| 145 | Markus Malo | Quantitative Characterization of Proximal … · Proximal Femur Using Pulse-Echo Ultrasound Measurements Osteoporosis is a common bone disease leading to increased

Publications of the University of Eastern Finland Dissertations in Forestry and Natural Sciences

Publications of the University of Eastern Finland

Dissertations in Forestry and Natural Sciences

isbn 978-952-61-1530-6

Markus Malo

Quantitative Characterization of Proximal Femur Using Pulse-Echo Ultrasound Measurements

Osteoporosis is a common bone

disease leading to increased fragility

and fracture probability. However,

only quarter of individuals suffering

from the disease have received a

diagnosis. For effective management

of the disease it would be highly

important to develop diagnostic

tools capable of mass screening of

the population at the basic level of

healthcare. In this thesis quantitative

pulse-echo ultrasound technique for

evaluation of proximal femur was

investigated and developed towards

this goal by means of numerical

modelling and in vitro and ex vivo

measurements.

dissertatio

ns | 145 | M

ar

ku

s Ma

lo | Q

ua

ntita

tive Ch

ara

cterizatio

n of P

roxim

al F

emu

r Usin

g Pu

lse-Ech

o U

ltraso

un

d...

Markus MaloQuantitative

Characterization of Proximal Femur Using Pulse-Echo Ultrasound

Measurements

ii

ii

ii

ii

MARKUS MALO

QuantitativeCharacterization of

Proximal Femur UsingPulse-Echo Ultrasound

Measurements

Publications of the University of Eastern FinlandDissertations in Forestry and Natural Sciences

No 145

Academic DissertationTo be presented by permission of the Faculty of Science and Forestry for publicexamination in the Auditorium SN200 in Snellmania Building at the University

of Eastern Finland, Kuopio, on September, 20, 2014, at 9 a.m.

Department of Applied Physics

ii

ii

ii

ii

Grano Oy

Kuopio, 2014

Editor: Prof. Pertti Pasanen, Prof. Pekka Kilpeläinen

Prof. Kai Peiponen, Prof. Matti Vornanen

Distribution:

University of Eastern Finland Library / Sales of publications

P.O. Box 107, FI-80101 Joensuu, Finland

tel. +385-50-3058396

http://www.uef.fi/kirjasto

ISBN: 978-952-61-1530-6 (printed)

ISBN: 978-952-61-1531-3 (PDF)

ISSNL: 1798-5668

ISSN: 1798-5668

ISSN: 1798-5676 (PDF)

ii

ii

ii

ii

Author’s address: University of Eastern FinlandDepartment of Applied PhysicsP.O.Box 162770211 KUOPIOFINLANDemail: [email protected]

Supervisors: Professor Juha Töyräs, Ph.D.University of Eastern FinlandDepartment of Applied Physicsemail: [email protected]

Professor Jukka Jurvelin , Ph.D.University of Eastern FinlandDepartment of Applied Physicsemail: [email protected]

Associate Professor Hanna Isaksson, Ph.D.University of Eastern FinlandDepartment of Applied Physics

Lund UniversityDepartment of Biomedical Engineeringemail: [email protected]

Reviewers: Professor Sulin Cheng, Ph.D.University of JyväskyläJyväskyläFinlandemail: [email protected]

Professor Mami Matsukawa, Ph.D.Doshisha UniversityLaboratory of Ultrasonic ElectronicsKyotanabe, KyotoJapanemail: [email protected]

Opponent: Frédéric Padilla, Ph.D.LabTau LaboratoryInserm UnitLyon, Franceemail: [email protected]

ii

ii

ii

ii

ii

ii

ii

ii

ABSTRACT

Bone fractures cause suffering and mortality, but also represent asignificant economic burden to the society due to lost work inputand direct hospital costs. Osteoporosis increases the risk of fracturethrough reduced bone mass and changes in the bone microstruc-ture. The current gold standard for diagnostics, i.e., dual energyx-ray absorptiometry (DXA), is available only in specialist health-care units and therefore, it is not used for mass screening. Thismight explain in part why as many as 75 % of osteoporotic patientslack proper diagnosis and medication. Thus, it is very importantto develop diagnostic tools capable of screening the population atthe basic level of healthcare. Quantitative ultrasound (QUS) de-vices for osteoporosis diagnostics have been available for decades.Their ability to predict fractures is similar to that of DXA. However,since it is used mainly for measurement of the extremities (heel,wrist), ultrasound has provided only a moderate estimate of boneproperties at the most important fracture sites. To further enhancethe fracture prediction, QUS measurements should be conducted atmore sensitive sites, e.g., at proximal femur. This site is covered bysoft tissues, which introduces errors into the ultrasound measure-ment.

In this thesis, changes in the cortical and trabecular bone tissueelastic coefficients and porosities during aging at the proximal fe-mur were assessed by means of ultrasound microscopy (Study I).Moreover, novel and traditional ultrasound backscatter parameterswere measured from intact proximal femur ex vivo and comparedwith the bone mineral density and trabecular structure (Study II).In addition, the ability of the dual frequency ultrasound (DFUS)technique to compensate for errors in bone ultrasound measure-ment due to soft tissue was evaluated. Furthermore, the error inDFUS arising from non-perpendicular ultrasound incidence at softtissue and soft tissue - bone interfaces was investigated with nu-merical simulations (Study III). Moreover, the effect of non-optimalfocusing to the soft tissue - bone interface on DFUS based correc-

ii

ii

ii

ii

tion for attenuation originating from the overlying soft tissues wasassessed with numerical simulations and in experimental measure-ments (Study IV).

The cortical bone tissue elastic coefficient and the porosity werefound to increase with age (R2 = 0.28 − 0.46, p < 0.05 − 0.01)(Study I). Furthermore, the elastic coefficient was significantly high-er (p < 0.05) in cortical bone than in trabecular bone and the valuevaried between anatomical locations (Study I). The backscatter pa-rameters measured ex vivo at the proximal femur were significantlycorrelated with bone mineral density (R2 = 0.45, p < 0.01) andtrabecular microstructure (R2 = 0.43, p < 0.01) (Study II). Non-optimal focusing of ultrasound to soft tissue - bone interface (StudyIV) and non-perpendicular ultrasound incidence at soft tissue andsoft tissue - bone interfaces (Study III) were found to induce signif-icant errors in the QUS measurements as well as in the DFUS es-timated soft tissue composition. However, in both studies (III andIV), with optimized ultrasound focusing and incidence at interfaces,the error in QUS parameters was significantly reduced by applyinginformation about the interfering layer thickness and composition,as obtained with the DFUS technique.

To conclude, measurement of QUS parameters from proximalfemur and minimization of the soft tissue related errors with theDFUS technique are possible and warranted. Since the porosity andelastic coefficient were found to vary with age, it would be highlyimportant to investigate these issues also in osteoporotic bones inorder to be able to distinguish between aging and osteoporosis re-lated changes in bone with QUS. The ex vivo measurements indi-cated that the QUS parameters were dependent on the bone mineraldensity and trabecular structure of intact proximal femurs. Thus,quantitative ultrasound backscatter measurements, supplementedwith DFUS correction for soft tissue induced errors, could enablescreening for osteoporosis at the level of basic healthcare. How-ever, to reach this, technical development, e.g., use of phased arraytechnique and extensive in vivo testing are needed.

ii

ii

ii

ii

National Library of Medicine Classification: QT 34.5, QT 36, WE 200,WE 250, WN 208

Medical Subject Headings: Bone and Bones; Bone Density; Femur;Hip Fractures; Osteoporotic Fractures; Osteoporosis/diagnosis; Ag-ing; Elastic Tissue; Biomechanical Phenomena; Elasticity; Porosity;Absorptiometry, Photon; Microscopy, Acoustic; Numerical Anal-ysis, Computer-Assisted; Computer Simulation; Ultrasonography;Ultrasonics

Luokitus: QT 34.5, QT 36, WE 200, WE 250, WN 208

Yleinen suomalainen asiasanasto: luu; luuntiheys; reisiluu; osteoporoosi- - diagnoosi; ultraääni; ultraäänitutkimus; kimmoisuus; huokoisuus;akustinen mikroskopia; ikääntyminen; röntgentutkimus; fotoniab-sorptiotekniikka; simulointi; numeeriset menetelmät

ii

ii

ii

ii

To Laura, Luukas and Eevi

ii

ii

ii

ii

Acknowledgements

What will remain from us?Hopefully children’s and warm memories to our loved ones.Possibly a lot of unfinished work.At the very least - 206 bones.

This study was carried out during the years 2009-2014 in the De-partment of Applied Physics at the University of Eastern Finland.

I would like to express my gratitude to my supervisors for theirprofessional guidance during this thesis project. I would like tothank my principal supervisor Juha Töyräs for the endless enthu-siasm towards research work and for the discussions we have hadboth about and out with the research topics. Moreover, I am grate-ful to my second supervisor Professor Jukka Jurvelin for the oppor-tunity to work in his top class research group, Biophysics of Boneand Cartilage (BBC). I also want to thank my third supervisor As-sociate Professor Hanna Isaksson for the tough questions and thepush to do critical thinking in research, but also for the understand-ing and soft values during the different phases of life during thisthesis.

I am grateful to the reviewers of this thesis, Professors SulinCheng and Mami Matsukawa, for their professional review and en-couraging comments. I would also like to thank Ewen MacDonaldfor linguistic review.

I would like to express my deepest gratitude to all of my co-authors for their significant contributions to the studies. Partic-ularly, I want to thank Janne Karjalainen, Sami Väänänen, JukkaLiukkonen, Katariina Nissinen, Mikael Turunen, Daniel Rohrbach,Kay Raum, Heikki Kröger, Xioyu Tong, Inari Tamminen, Ossi Riekki-nen, Antti Aula, Mikko Nissi, Erna Kaleva, Jari Rautiainen, Matti Ti-monen, Juuso Honkanen, Tuomo Silvast, Simo Saarakkala, TuomasViren, Harri Kokkonen, Roope Lasanen, Chibuzor Eneh, Xiaowei

ii

ii

ii

ii

Ojanen, Cristina Florea and Viktoria Prantner for the fruitful dis-cussions and their collaboration. Naturally I want to thank every-one working under the BBC group umbrella. It has been a pleasureand a privilege to work in such a stimulating atmosphere. More-over, I want to thank people from SIB labs, Arto Koistinen, RitvaSavolainen and Juhani Hakala, for the help and guidance in thesample preparation and the personnel from the Department of Ap-plied Physics, Jukka Laakkonen, Aimo Tiihonen, Tarja Holopainenand Heikki Väisänen for all kinds of technical support.

I also want to express my gratitude to all my relatives andfriends for giving the support and something else to think outsideof the work during the past years.

For financial support the strategic funding of University of East-ern Finland, Kuopio University Hospital (EVO grants), Kuopio Uni-versity foundation, Emil Aaltonen foundation, the Finnish Founda-tion for Technology Promotion and National Doctoral Programmeof Musculoskeletal Disorders and Biomaterials (TBDP) are acknowl-edged.

Finally, I am grateful to my parents, Irja and Ilkka, and mybrother Petri and his family for their continuous support, encour-agement and love throughout my life. I owe my deepest gratitudeto my beloved wife Laura and our two miracles Luukas and Eevi.Thank you Laura for your endless love, encouragement, supportand understanding during all these years.

Kuopio, August 2014

Markus Malo

ii

ii

ii

ii

ABBREVIATIONS

3D three-dimensionalA/D analog to digitalAIB apparent integrated backscatterANOVA analysis of varianceBMD bone mineral densityBUA broadband ultrasound attenuationCT X-ray computed tomographyCV coefficient of variationDFUS dual-frequency ultrasoundDOF depth of fieldDXA dual energy X-ray absorptionFDTD finite difference time domainFemUS femur ultrasound scannerFSAB frequency slope of apparent backscatterFWHM full width at half maximumIRC integrated reflection coefficientMBD mean of the backscatter differenceNAHNES National Health and Nutrition Examination SurveyPBS phosphate buffered salinePDE partial differential equationsPE pulse-echoPMMA polymethylmethacrylatePZT lead zirconate titanateQUS quantitative ultrasoundRANK Receptor Activator for Nuclear Factor k BRMS root mean squareSAM scanning acoustic microscopySD standard deviationSOS speed of soundTOF time of flightTPX polymethylpenteneTSAB time slope of apparent integrated backscatterTT through-transmission

ii

ii

ii

ii

WHO World Health OrganizationµCT micro computed tomography

ii

ii

ii

ii

SYMBOLS AND NOTATIONS

AA average attenuationA amplitudeA(f) amplitude spectrumAr areaα attenuation coefficientb medium dependent coefficientc speed of soundcp phase velocitycl longitudinal sound velocitycs shear sound velocitycm speed of sound in mediumcw speed of sound in waterCii elastic stiffness coefficient to direction iid distanceDi diameter of transducerE Young’s modulusf frequencyH the term including ultrasound reflections from different

surfacesI acoustic intensityki correction factor for compensation of ultrasound reflection

at the adipose-lean interfaceK bulk modulusλw wavelengthk wavenumberZ acoustic impedanceρ mass densityν Poisson’s ratiop sound wave pressure or statistical differences tissue interfacevu particle velocity

ii

ii

ii

ii

θ angleR correlation coefficientRC reflection coefficient (intensity)TR transmission coefficient (intensity)d distancet timem medium dependent coefficientσ stressϵ strainλ first Lamé constantµ second Lamé constantF force or value of F-test (statistics)l lengthFl focal lengthRl radius of lens curvatureDOF depth of fieldx thicknessc speed of soundn number of samplesSMI structural model indexTb.N trabecular numberTb.Sp trabecular separationTb.Th trabecular thickness∆ f (x) forward difference form∇ f (x) backward difference formδ f (x) central difference form∂ partial difference operator∇ gradient operator∇· divergence operatoru particle displacementω angular frequency of the waveϕ phase angle

ii

ii

ii

ii

LIST OF PUBLICATIONS

This thesis is based on the following original articles referred to bytheir Roman numerals:

I M. K. H. Malo, D. Rohrbach, H. Isaksson, J. Töyräs,J. S. Jurvelin, I. S. Tamminen, H. Kröger, K. Raum, “Longi-tudinal elastic properties and porosity of cortical bone tissuevary with age in human proximal femur,” Bone 53(2), 451-8(2013).

II M. K. H. Malo, J. Töyräs, J. P. Karjalainen, H. Isaksson,O. Riekkinen and J.S. Jurvelin, “Ultrasound backscatter mea-surements of intact human proximal femurs – relationshipsof ultrasound parameters with tissue structure and mineraldensity,” Bone 64, 240-5 (2014).

III M. K. H. Malo, J. P. Karjalainen, H. Isaksson, O. Riekkinen,J. S. Jurvelin, J. Töyräs, “Numerical analysis of uncertaintiesin dual frequency bone ultrasound technique,” Ultrasound inMedicine and Biology 36(2), 288-94 (2010).

IV M. K. H. Malo, J. P. Karjalainen, O. Riekkinen, H. Isaksson,J. S. Jurvelin, J. Töyräs, “Effects of non-optimal focusing ondual-frequency ultrasound measurements of bone,” IEEE Trans-actions on Ultrasonics, Ferroelectrics, and Frequency Control 58(6),1182-8 (2011).

The original articles have been reproduced with kind permission ofthe copyright holders.

ii

ii

ii

ii

ii

ii

ii

ii

AUTHOR’S CONTRIBUTION

The publications selected to this dissertation are original researchpapers on bone ultrasound measurements and experimental andnumerical evaluation of the error sources with the dual frequencyultrasound technique. The author has contributed to the study de-sign and carried out all measurements and analyses, except for partof the micro-computed tomography imaging in study II. The au-thor has written the manuscripts of studies I-IV. In all papers theco-operation with the co-authors has been significant.

ii

ii

ii

ii

ii

ii

ii

ii

Contents

1 INTRODUCTION 1

2 BONE 52.1 Skeleton . . . . . . . . . . . . . . . . . . . . . . . . . . 52.2 Structure and composition . . . . . . . . . . . . . . . . 52.3 Remodeling . . . . . . . . . . . . . . . . . . . . . . . . 82.4 Aging and osteoporosis . . . . . . . . . . . . . . . . . 10

3 X-RAY ASSESSMENT OF BONE 133.1 Dual-energy X-ray absorptiometry . . . . . . . . . . . 133.2 Computed tomography . . . . . . . . . . . . . . . . . 14

4 ULTRASOUND ASSESSMENT OF BONE 174.1 Basic physics of ultrasound . . . . . . . . . . . . . . . 174.2 Acoustic properties of tissues . . . . . . . . . . . . . . 184.3 Ultrasound scattering and reflection . . . . . . . . . . 214.4 Ultrasound absorption . . . . . . . . . . . . . . . . . . 234.5 Ultrasound applications . . . . . . . . . . . . . . . . . 244.6 Scanning acoustic microscopy . . . . . . . . . . . . . . 284.7 Dual frequency ultrasound (DFUS) technique . . . . 314.8 Numerical modelling of acoustic wave propagation . 35

5 AIMS OF THE PRESENT STUDY 39

6 MATERIALS AND METHODS 416.1 Materials . . . . . . . . . . . . . . . . . . . . . . . . . . 416.2 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . 43

6.2.1 Basic characteristics of bone . . . . . . . . . . . 436.2.2 Ultrasound experiments . . . . . . . . . . . . . 446.2.3 Numerical simulations . . . . . . . . . . . . . . 476.2.4 Statistical analysis . . . . . . . . . . . . . . . . 51

ii

ii

ii

ii

7 RESULTS 537.1 Elastic coefficient and porosity in human cortical bone 537.2 Ultrasound backscatter measurement of proximal fe-

mur ex vivo . . . . . . . . . . . . . . . . . . . . . . . . . 557.3 Age-related changes in bone . . . . . . . . . . . . . . . 577.4 Performance of the dual frequency ultra-

sound technique . . . . . . . . . . . . . . . . . . . . . . 59

8 DISCUSSION 658.1 Elastic coefficients and porosity of bone tissue vary

during aging . . . . . . . . . . . . . . . . . . . . . . . . 668.2 Ultrasound backscatter in proximal femur is related

to bone density and microstructure . . . . . . . . . . . 678.3 Evaluation of the dual frequency ultrasound technique 69

9 SUMMARY AND CONCLUSIONS 71

BIBLIOGRAPHY 74

ii

ii

ii

ii

1 Introduction

The number of elderly individuals is growing as the average life ex-pectancy increases. There are several age-related changes occurringin the human body, for example, the human skeleton goes throughvarious developmental and degenerative phases. The skeleton reach-es full maturity and peak bone mass at the age of 25. Thereafter,skeletal degeneration begins, i.e., there is a reduction in bone massand a deterioration in the bone quality [1–8]. The term bone qual-ity reflects many factors, e.g., bone architecture, turnover, damageaccumulation (e.g., microfractures) and mineralization [9].

These changes reduce the mechanical competence of the wholebone. Thus, with age the fracture probability increases [10]. Bonefractures not only lead to increased mortality rate but they are alsoresponsible for cause significant financial expenditures via both di-rect hospital costs and in the form of lost work capacity [10, 11, 11–13]. Thus, prevention of fractures is highly important for both theindividual and society as a whole.

Osteoporosis is the most common skeletal disease in the elderly.In global terms, it is estimated that about 200 million people haveosteoporosis, 27.5 million of these live in Europe [13, 14]. Osteo-porosis is a systemic skeletal disease characterized by decreasedbone mass and deterioration of the bone microstructure. It resultsin increased bone fragility and on elevated fracture risk [15,16]. Theintrinsic properties of bone, e.g., bone material properties, shapeand architecture, define the capability of bone to resist fractur-ing [17–19]. However, the fracture risk is also affected by manyexternal factors which can influence the susceptibility to fall, e.g.,environmental conditions, individual lifestyle, vision, balance, re-action time and muscle strength [20, 21]. Therefore, the accurateestimation of the fracture risk is a challenging problem.

The gold standard for osteoporosis diagnosis is the measure-ment of bone mineral density (BMD) with dual energy absorptiom-

ii

ii

ii

ii

Markus Malo: Quantitative Characterization of Proximal Femur UsingUltrasound Pulse-Echo Measurements

etry (DXA) [16]. The DXA devices used for the osteoporosis diag-nostics are generally available only in specialized healthcare centersand are not used for mass screening of the population [22]. Thismight explain in part why as many as 75 % of the osteoporoticpatients are not properly diagnosed and thus fail to receive appro-priate therapy [23]. Thus, it would be very important if one coulddevelop diagnostic tools capable of screening the population at thelevel of basic healthcare.

Quantitative ultrasound (QUS) has been proposed as being anon-ionizing option for screening for osteoporosis [24–26]. QUS issensitive to bone elastic properties, density, structure and also itcan estimate the inorganic phase that cannot be evaluated by X-raymethods [27]. Although peripheral ultrasound measurement de-vices have been on the market for decades, they have not achievedany major clinical breakthroughs [28]. There are several reasons(e.g. measurement of extremities) why ultrasound seems to obtainonly a moderate estimation of bone properties at the axial skele-ton [24, 29]. If one wishes to enhance the fracture prediction atthe most serious fracture sites, e.g., the proximal femur, site spe-cific measurements are needed [30–32]. However, bones in the ax-ial skeleton are covered with a substantial soft tissue layer and itsthickness and composition vary from patient to patient introducingerrors into QUS measurements.

The speed of sound in bone depends on its physical density andelastic properties. For example, when evaluating the cortical bonethickness, it is important to know the age dependent variation offactors affecting the speed of sound. With aging, the trabecularmicrostructure within the bone undergoes changes. Common find-ings include a reduction in the trabecular network connectivity andthe number of trabeculaes, but also thinning and a change in shapeof single trabeculae. Further, these changes may be measured invivo by dual energy absorptiometry as a decrease in the areal bonemineral density. However, the potential of ultrasound backscatterto detect these changes has not been investigated in intact proximalfemur ex vivo. It would be very important to determine whether the

2 Dissertations in Forestry and Natural Sciences No 145

ii

ii

ii

ii

Introduction

trabecular microstructure and bone mineral density can be evalu-ated at the most serious fracture site, i.e., proximal femur with ul-trasound backscatter measurements before undertaking extensivein vivo evaluation of this technique.

A dual frequency ultrasound (DFUS) technique has been de-veloped for the measurement of soft tissue thickness and compo-sition [33]. The layered and non-homogenous structure of soft tis-sue causes attenuation and scattering thus distorting the ultrasoundsignal measured from bone. If one incorporates information aboutthe soft tissue covering the bones, the error caused by the soft tissuein the bone QUS parameters may be minimized. This technique hasbeen tested in vitro and applied in vivo [33, 34]. However, its sen-sitivity to the error sources that are present in vivo have not beenfully characterized. Since bones are often located under a thicklayer of soft tissue, focused ultrasound transducers need to be ap-plied to maximize the signal to noise ratio. Unfortunately, due toindividual variation in soft tissue thickness, it is not always possibleto conduct the measurement at the optimal focal distance from thebone surface. Moreover, non-perpendicular ultrasound incidenceat tissue interfaces cause distortion and refraction to the propagat-ing ultrasound pulse. The non-perpendicular ultrasound incidenceat tissue interfaces and non-optimal focusing may introduce errorswhen analyzing the thickness and composition of the soft tissuewith the DFUS method. Thus, clarification and the elimination ofthese sources of error would be very advantageous.

The present study aims to fill these gaps in our knowledge.The changes in the porosity and elastic coefficient of cross-sectionalbone samples obtained from the femoral neck and shaft have beenevaluated using scanning acoustic microscopy (SAM). Furthermore,the sensitivity of ultrasound backscatter in the estimation of bonemineral density and trabecular structure has been determined inthe proximal femur ex vivo. Finally, numerical simulations and ex-perimental measurements have been applied to examine the errorsources related to application of the DFUS method.

Dissertations in Forestry and Natural Sciences No 145 3

ii

ii

ii

ii



ii

ii

ii

ii

2 Bone

2.1 SKELETON

The human skeleton can be divided into the axial skeleton, includ-ing the vertebrae and pelvis, and the appendicular skeleton, includ-ing all the long bones. The skeleton consists of four types of bones:long bones, e.g., tibia and radius, short bones, e.g., phalanges, flatbones, e.g., ribs, and irregular bones, e.g., vertebrae. Bones pro-vide mechanical support for the body and enable locomotion. Theyfunction as levers and transform the forces from the muscles intomovements. Bones protect our vital organs, e.g., the brain, but theyalso store and release minerals such as calcium and phosphorus.Moreover, the inside of bones is the location of red or yellow bonemarrow. Red marrow produces red and white blood cells, andplatelets, hence it has a key role in hematology and the immunesystem. Yellow marrow primarily acts as a fat storage site [35–37].Skeletal bone mass increases rapidly during adolescence to reachits peak value when the individual is around 25 years of age [1, 3].Thereafter, the bone mass starts to slowly decline. In women, afterthe menopause, the reduction in bone mass occurs rapidly [4, 38].

2.2 STRUCTURE AND COMPOSITION

Structure

Bone tissue can be divided into cortical and trabecular bone basedon its structure at the macroscopic level. The division can also bebased on the maturity of the bone tissue. Woven and lamellar bonerepresent newly developed and mature bone, respectively. In wo-ven bone, the collagen fibril network is randomly organized andthe osteocyte and water contents are higher than in fully maturebone [39, 40]. Woven bone undergoes a rapid rate of depositionand turnover, and its mineralization pattern is irregular. Duringmaturation of lamellar bone, its structure including the collagen


ii

ii

ii

ii


fibers and hydroxyapatite crystals, becomes organized. Lamellarbone exists both in trabecular and cortical bone. Lamellar bone isstiffer than woven bone, due to its highly organized structure. Fur-thermore, the mechanical properties of woven bone are isotropicwhereas lamellar bone is mechanically anisotropic [35]. Accordingto Wolff’s Law, the bone adapts its internal architecture and externalshape to resist forces in the main mechanical loading direction [41].A schematic illustration of bone structure is presented in Figure 2.1.

Cortical bone

Cortical bone is compact (approximately 5 to 10 % porosity) anddense (1600-2000 kg/m3) and it forms the outer layer of the bones[42, 43]. The human skeleton is mainly (80 - 90 %) composed ofcortical bone [35, 44, 45]. The cortex is thick especially in the dia-physis of the long bones, providing maximum resistance to torsionand bending. In the epiphyisis, the thin cortex is supported by theunderlying trabecular structure enabling high deformation duringloading. Cortical bone is formed of packed osteons and interstitialtissue. The osteons are connected by the Haversian system [36, 37].In comparison to the porous trabecular bone, cortical bone has aslow turnover rate and lower metabolism.


ii

ii

ii

ii

Bone

Figure 2.1: A schematic illustration of cortical and trabecular bone structure, showing thelamellar structure, Haversian system, peri-, and endosteum. (Figure is modified from [35]).

Trabecular bone

Trabecular bone consists of a network of trabeculae and it is highlyporous. The typical diameter of a trabeculae is 100 to 200 microm-eters [15]. The volumetric density of trabecular bone is much lowerthan that of cortical bone. Thus, the surface area of the trabecularbone is about twenty times higher than that of cortical bone of a


ii

ii

ii

ii


similar mass. However, the density of the calcified matrix is ap-proximately the same in cortical and trabecular bone. Trabecularbone is present particularly in the vertebrae, and at the ends of thelong bones. The volume between the trabeculae is filled with bonemarrow. Since the cells in the trabecular bone are located on thesurface of the trabeculae, they are in proximity or in direct con-tact with the bone marrow, blood vessels and cells. This togetherwith high surface-to-volume area facilitates a high rate of metabolicactivity and good sensitivity to remodeling under mechanical load-ing [4, 35].

Composition

Bone is a composite material consisting of an organic and an in-organic phase. Approximately 70 % of the weight of the bone ismineralized material, i.e., crystalline hydroxyapatite which makesthe bone hard and stiff, and enables it to resist compression. Be-tween 5 % to 8 % of the weight of the bone is water, with theremaining part being organic material consisting mostly of colla-gen type I (90 %) and a variety of non-collagenous proteins [46].The organic component provides tensile strength and elasticity tothe bone and enables the bone to deform and resist stretching andtwisting [47]. The bones store the majority of the minerals in hu-man, i.e., they represent a store from which calcium, phosphorus,sodium and magnesium ions can be released into the extracellularfluid [35].

2.3 REMODELING

The skeleton undergoes remodeling throughout the lifespan. Infact, most of the skeletal system is replaced approximately every10 years [48]. This remodeling is achieved by resorption of bonematrix by osteoclasts and its replacement by osteoblasts [49].

Bone resorption by osteoclasts

Osteoclasts dissolve the old bone by secreting acids and enzymes.


ii

ii

ii

ii

Bone

Osteoclasts are formed from the same precursor cells as the whiteblood cells [48]. Osteoclasts become activated when Receptor Ac-tivator for Nuclear Factor κ B (RANK) receptors (in osteoclast pre-cursors) are stimulated by RANK ligand (secreted by osteoblasts).Furthermore, osteoprotegerin binds RANK ligand and in this wayit regulates osteoclast activation [50]. When an osteoclast has fin-ished bone resorption, it undergoes apoptosis.

Bone formation by osteoblasts

When old bone has been removed by osteoclasts, the empty spaceis replaced with new bone produced by osteoblasts. Osteoblastsare formed from marrow precursor cells, which are also capable ofdifferentiating into fat cells. The osteoblasts produce proteins thatform the organic matrix and control the mineralization. Osteoblastreceptors are sensitive to several hormones, e.g., estrogen, parathy-roid hormone and vitamin D. Moreover, they communicate withother cells, e.g., osteoblasts secreting RANK ligand [35, 48].

Osteocytes and lining cells

When osteoblasts have finished forming new bone, a part of thembecome trapped in the matrix where they differentiate into osteo-cytes. The cells remaining at the surface of the matrix differenti-ate into bone lining cells while the remaining osteoblasts undergoapoptosis. Osteocytes are arranged circularly around osteon andare connected through canaliculi which have diameters of a fewmicrometers [51] (Figure 2.1). They are able to signal and activateboth lining cells and osteoblasts. Lining cells are flat shaped cellsthat cover the bone surfaces. The lining cells possess receptors forhormones and growth factors and may become activated to initi-ate bone remodeling when necessary [52]. The remodeling processcomplies with Wollf’s Law and, thus, the bone adapts to the domi-nant direction of mechanical loading.


ii

ii

ii

ii


2.4 AGING AND OSTEOPOROSIS

During adolescence, the activity of osteoblasts is greater than thatof the osteoclasts, but with increasing age, the osteoblast activitybecomes reduced. This may lead to a deterioration of the bonemicroarchitecture, i.e., an increase in porosity and a thinning ofcortex and trabeculae, and changes in the bone material proper-ties [4, 8, 53–56]. Thus, with aging, the bone structure weakens,which reduces the mechanical strength of the bone and increasesthe fracture risk [57, 58].

Figure 2.2: Two-dimensional µCT section of normal and osteoporotic trabecular bone. Thewhite color illustrates the trabecular bone structure of a cylindrical bone sample extractedfrom a human proximal tibia.

Osteoporosis means - porous bone in latin. The osteoporoticchanges in the skeleton appear as a reduction in the bone mineraldensity in the whole skeleton, i.e., a deterioration in the bone macro-, and microarchitecture, such as an, increase in cortical porosity,a thinning of the trabeculae and cortices, and a decreased num-ber of trabeculaes (Figure 2.2) [59]. These changes increase therisk of fracture since they weaken the structure of the remainingbone [4, 31, 32, 60, 61]. As the trabecular bone has a large surface tovolume ratio, it is a sensitive reflection of the appearance of osteo-porosis. There may be various reasons for the skeletal degeneration,but the overall result is that the osteoclasts remove more bone than


ii

ii

ii

ii

Bone

the osteoblasts can produce. Approximately 40 % of women and15 % of men over 50 years of age will suffer one or more fragilityfractures during their remaining lifetime [38]. On average, the an-nual declines from the peak bone mass in areal bone mineral den-sity at the femoral neck are 0.4 percent for men and 0.5 percent forwomen [62].

Diagnostics of osteoporosis

The diagnosis of osteoporosis is based on DXA measurement of thebone mineral density in the proximal femur or spine [16]. Accord-ing to the World Health Organization (WHO) and the InternationalOsteoporosis Foundation recommendations, osteoporosis diagnosisis based on the T-score [63,64]. The T-score is defined as the numberof standard deviations (SD) above or below the mean bone mineraldensity of a young adult reference population. The age-adjustedrelative increase in fracture risk is approximately doubled for eachSD decrease in BMD [31, 32, 61]. A hip BMD value greater than 1SD below the young adult reference population mean (T-score ≥–1) is considered as normal. A hip BMD value between 1 SD and2.5 SD below the young adult reference population mean (T-score <–1 and > –2.5) is referred to as being osteopenic, i.e. low bone mass.A hip BMD value of 2.5 SD or more below the young adult refer-ence population mean (T-score ≤ –2.5) is diagnosed as osteoporo-sis, and in addition if there have been one or more fragility frac-tures, then the diagnosis is severe osteoporosis, i.e., established os-teoporosis [16]. Currently the recommended reference database isthe National Health and Nutrition Examination Survey (NHANES)III which contains femoral neck measurements from 20-29 year oldwomen. Although there has been discussion and conflicting reportsabout whether the same cut-off values for osteoporosis may be ap-plied for the male population, the present recommendation is touse the same database and criteria for men and women [65].

Screening of osteoporosis with DXA in population is not recom-mended and instead, screening should be targeted to those patientsbelonging to a risk group [22, 61, 66]. In order to identify these


ii

ii

ii

ii


patients, a Fracture Risk Assessment Tool FRAX) has been devel-oped by WHO [67]. The FRAX tool is based on individual patientmodels that integrate the risks associated with several clinical fac-tors. The FRAX tool can be used with the national osteoporosisguidelines by which patients can be categorized into three treat-ment groups: lifestyle advice and reassure, measure BMD and treat.These thresholds are provided only to support clinicians when con-sidering which patients would benefit from further treatment.

The best prediction of fracture seems to be obtained by site spe-cific measurements, e.g., the best prediction for hip fracture is ob-tained by the DXA measurements of BMD from the proximal femur.Measurement of BMD with DXA is quite reproducible (coefficientof variation 2 %). This is very important when evaluating the ef-fectiveness of the medical treatment or conducting follow-up stud-ies [68]. Although DXA is a reproducible tool for measuring theareal BMD, it only provides indirect information on the bone me-chanical properties and gives no information about the microarchi-tecture or the organic composition of the bone. Moreover, changesin the composition of the overlying soft tissue affect the determinedvalues of BMD implying that the method for accounting for the softtissue related factors is not optimal [69]. Furthermore, it has beenclaimed that the variation in bone marrow composition may intro-duce error to the value of BMD measured [70].


ii

ii

ii

ii

3 X-ray assessment of bone

Since the discovery of X-rays in 1895, X-ray imaging has been widelyused in clinical diagnostics. Plain X-ray imaging can be used to de-tect bone fractures and subchondral sclerosis related to osteoarthri-tis, and the dual energy X-ray absorptiometry (DXA) is used for thediagnostics of osteoporosis.

3.1 DUAL-ENERGY X-RAY ABSORPTIOMETRY

Dual-energy X-ray absorptiometry (DXA) measures the areal bonemineral density (aBMD, g/cm2). There are two main types of DXAdevices. There are peripheral devices which can be used for themeasurement of the heel or the radius, and there are whole bodyDXA scanners, which are more commonly used. In whole bodyscanners, the X-ray source, the collimator and the detector are im-plemented into a C-arm. The DXA devices are based either onpencil or fan beam technology. The ’Dual Energy’ refers to fact thatX-ray radiation with two different energy levels, such as, 38 and 70keV is being used [71]. As the patient is scanned with two differentenergy levels, two different attenuation profiles are obtained. Theareal BMD, e.g., from the proximal femur or lumber vertebra, canbe evaluated by determining the soft tissue thickness and composi-tion adjacent to the bone and compensating for its effect in the bonemeasurement. The typical regions of interest in the proximal femurare illustrated in Figure 3.1.


ii

ii

ii

ii


Figure 3.1: A screenshot from BMD measured by DXA in the proximal femur ex vivo.The left side of the figure shows the analyzed regions and the shape of the femur. The rightside of the figure shows an illustration of the average BMD in females as a function ofage (cyan represents ± 1 SD limits) in which the black dot corresponds to the measuredfemoral neck BMD value.

The radiation dose obtained from a whole body DXA examina-tion is small (4 - 30 µSv), corresponding to approximately four dailyamount of natural radiation [71, 72]. The duration for a total bodyscan with a fan and pencil beam technology is less than 10 and 20minutes, respectively [71], whereas a hip measurement only takesless than a minute.

3.2 COMPUTED TOMOGRAPHY

Computed topography (CT) is based on X-ray attenuation projec-tion images of an object from multiple angles of rotation (rotation ofat least 180 degrees). These images are further processed by math-ematical algorithms to provide a three dimensional (3D) graphicreconstruction of the imaged object. With clinical CT devices, thetypical isotropic voxel size in the images varies from 0.5 x 0.5 x


ii

ii

ii

ii

X-ray assessment of bone

0.5 mm3 to 5 x 5 x 5 mm3. By increasing the radiation dose, theisotropic voxel size can be decreased to 300 x 300 x 300 µm3. Witha dedicated peripheral CT one can achieve, an isotropic resolutionof 82 x 82 x 82 µm3 (Scanco XtremeCT) [73]. However, the imag-ing time is approximately 3 minutes per centimeter. With the latestdevelopments in cone beam techniques, the new peripheral CT de-vice (Planmed Verity) is capable of capturing a 130 mm x 160 mmfield of view with a 200 x 200 x 200 µm3 isotropic voxel size in 17seconds [74]. For comparison, the isotropic resolution with a mod-ern desktop µCT is around 1 - 200 µm3. The field of view in µCTdevices is limited to the millimeter scale and, thus, they are capableof imaging only small animals and in vitro samples. Furthermore,the resolution comes with the cost of increased radiation dose andprolonged imaging time. Thus, it can take from tens of minutes tohours if one wishes to obtain high resolution images with a µCT.However, due to the nature of the objects being imaged, this isusually not a problem. In order to determine trabecular structurereliably, a resolution of at least 28 µm is desirable [75]. At the mo-ment, there is no clinical device capable of reliably estimating thetrabecular microstructure in vivo.

The high resolution 3D CT-images can be utilized to calculatemany geometrical parameters describing the imaged structure. Typ-ically from cortical bone, its thickness and volumetric porosity areevaluated, whereas from trabecular bone other parameters, e.g.,bone volume to total volume (BV/TV) and trabecular thickness,shape and connectivity are evaluated [6, 8, 55, 76, 77]. While DXAprovides information on areal BMD, CT determines the volumetricbone mineral density (vBMD) and the 3D geometry of the bone [72].This information is extremely valuable since it has been shown toprovide a more reliable estimation of the fracture risk than BMDmeasurements made by DXA alone [19, 78, 79]. However, the radi-ation dose utilized in a CT measurement may be a hundred timeshigher than that needed for DXA measurements, which limits theformer’s use in screening for osteoporosis. It has been proposedthat DXA measurement alone would allow estimation of the bone’s


ii

ii

ii

ii


3D shape with the BMD value and some index of its mechanicalstrength [80]. However, this approach requires patient specific fi-nite element models, but would provide information on the fractureload and what factors affect the fracture susceptibility [81].

10mm

Figure 3.2: A three-dimensional reconstruction of a cubioid bone sample harvested fromthe trochanter of a male cadaver. In the sample, dense cortical bone lies over the poroustrabecular bone. The isotropic voxel size of the µCT imaging was 34 x 34 x 34 µm3.


ii

ii

ii

ii

4 Ultrasound assessment ofbone

Quantitative ultrasound (QUS) techniques have been used for theassessment of bone for decades [28]. The ability of QUS to pre-dict fractures is approximately the same as can be achieved withDXA [82–85]. Ultrasound is sensitive to many bone - related prop-erties, i.e., density, structure, composition and mechanical proper-ties, and it provides also information on bone organic phase [86–91].However, the interaction between the bone matrix and ultrasoundwave field is complex and, therefore, it is challenging to relate ul-trasound parameter values to bone properties. Thus, for decadesmajor efforts have been made to gain a comprehensive understand-ing of the issues related to bone ultrasound measurements [28, 92].QUS does hold the potential to become an alternative to X-ray basedtechniques for mass screening of the population to identify those in-dividuals with an elevated risk of suffering fractures [25, 84, 85, 93].Moreover, ultrasound instrumentation can be manufactured to havea small size and the devices themselves are inexpensive. However,currently the clinical use of ultrasound is limited. This is partlybecause the diagnostic criteria for osteoporosis, as defined by theWorld Health Organization (WHO), is based on areal bone mineraldensity obtained by DXA [63, 64].

4.1 BASIC PHYSICS OF ULTRASOUND

Ultrasound is defined as a propagating mechanical wave with afrequency higher than can be detected by the human ear (20 kHz).The wave propagation is based on displacement of medium parti-cles from their resting positions which induces displacements of theneighboring particles. When the particle becomes displaced from


ii

ii

ii

ii


the equilibrium position, the restoration forces together with theinertia of the particles results in oscillation. The fundamental equa-tions describing wave propagation are presented in Table 4.1. In thepresent thesis, only longitudinal waves were investigated. However,there are also several other wave modes, e.g., transverse (shear) andsurface (Rayleigh) waves. The longitudinal wave can propagate inall types of material (solids, liquids and gases), since the energy istransferred through compression and expansion of the medium. Intransverse waves, the displacement of particles occurs in the direc-tion perpendicular to the direction of the wave propagation. Thesurface wave is a combination of the longitudinal and transversewaves, resulting in an elliptic orbit in particle displacement [94].

4.2 ACOUSTIC PROPERTIES OF TISSUES

All materials have characteristic acoustic properties, which maybe defined by the acoustic impedance and attenuation coefficient.Acoustic impedance is dependent of the volumetric mass densityof the material and the speed of sound. Speed of sound is de-pendent on Young’s modulus, Poisson’s ratio and volumetric massdensity, which are temperature dependent (Table 4.2). The acousticproperties of bone tissue depend on its collagen and mineral con-tents [88]. Moreover, the bone structure affects the acoustic prop-erties. In highly porous materials, e.g., trabecular bone, a slow andfast wave can be detected [98–104]. The fast wave is consideredto result from the displacement component in-phase in the min-eralized tissue and marrow, whereas the slow wave is a result ofthe components oscillating out of phase [99]. Furthermore, for dis-persive materials, both group and phase velocities may be deter-mined [105–109]. Phase velocity is defined individually for eachfrequency component whereas the group velocity is described forthe envelope of the wave [110]. Attenuation quantifies the lossof energy as the wave propagates through the medium. Attenua-tion results mainly from scattering, reflection, absorption and beam


ii

ii

ii

ii

Ultrasound assessment of bone

spreading [95]. These phenomena will be elaborated in the follow-ing sections.


ii

ii

ii

ii


Table 4.1: Basic equations describing the propagation and behavior of planar sound wavesin medium and at the interface of mediums [27, 95–97]

Parameter Equation

Particle displacement [m] u = u0sin(ωt − ϕ)

Angular frequency of particle [rad/s] ω = 2π f

Wavelength [m] λw =cpf = cpT

Wavenumber [m-1] k = 2πλw

= ωcp

Acoustic impedance [rayl] Z = ρcl

Longitudinal sound velocity in isotropic elas-tic solid [m/s]

cl =

√E(1−ν)

ρ(1+ν)(1−2ν)

Longitudinal sound velocity in fluids [m/s] cl =√

Kaρ

Shear sound velocity in isotropic elastic solid[m/s]

cs =√

Eρ(1+ν)

Acoustic intensity [W/m2] I = p2

2Z

Sound wave pressure for plain waves [Pa] p = ρclcu

Snell’s law [-] sinθ1c1

= sinθ2c2

= sinθ3c3

Reflection coefficient (intensity) [-] RC = ( Z2cosθ1−Z1cosθ2Z2cosθ1+Z1cosθ2

)2

Transmission coefficient (intensity) [-] TC = 4Z1Z2cos2θ1(Z2cosθ1+Z1cosθ2)2

Attenuation law [-] p(d) = p0e−αdandI = I0e−2αd

Attenuation coefficient [dB/cm] α = b f m

f = frequency, cp = phase velocity, T = period, ϕ = phase angle, E = Young’smodulus, ν = Poisson’s ratio, ρ = density, Ka = adiabatic bulk modulus, cu =particle velocity, d = distance, t = time, b = medium dependent coefficient andm = medium dependent coefficient. θ1 and θ2 are the angles of the incidenceand refraction, respectively. Subscripts 1, 2, and 3 refer to the first and secondmedium, and shear wave, and l and s to the longitudinal and shear soundvelocity.


ii

ii

ii

ii


Table 4.2: Material properties can be described with the following equations [27, 111, 112]

Parameter Equation

Elastic stiffness coefficient (Pa) (orthotropic) Cii = ρc2li

Young’s modulus (Pa) E =σ

ϵ=

µ(3λ + 2µ)

λ + µ

Bulk modulus (Pa) K =E

3(1 − 2ν)= λ +

23

µ

Poisson’s ratio (-) ν = − ϵlateralϵlongitudinal

=λ

2(λ + µ)

Stress (Pa) σ =F

Ar

Strain (-) ϵ =∆ll0

First Lamé constant (Pa) λ =Eν

(1 + ν)(1 − 2ν)

Second Lamé constant (shear modulus) (Pa) µ =E

2(1 + ν)

C = Elastic stiffness coefficient, ρ = mass density, c = speed of sound, E =Young’s modulus, σ = stress, ϵ = strain, λ = First Lamé constant, µ = SecondLamé constant, K = Bulk modulus, ν = Poisson’s ratio, F = force, Ar = area,∆l = change is the length and l0 = original length. Subscripts l and i refer tolongitudinal and in the direction of orthogonal symmetry, respectively.

4.3 ULTRASOUND SCATTERING AND REFLECTION

Scattering arises when the proceeding wave encounters inhomo-geneity in the medium density or elastic properties, i.e., a scatterer[96]. Furthermore, roughness of the interfaces induces scattering.The wavelength and the dimensions of the scatterer determine thescattering phenomenon. If the scatterers are much smaller than thewavelength, the phenomenon is called Rayleigh scattering. When


ii

ii

ii

ii


the scatterer is smaller than the wavelength, the scattering patterncan be described as monopole radiation or dipole radiation in thecase of changes in the elastic properties or volumetric mass den-sity, respectively. When the wavelength and the size of the scattererare similar, the scattering pattern becomes more complex [113]. Inthe case of trabecular bone, the scattering has been estimated byFaran scattering and weak scattering models [113, 114]. Typicallyin Faran solution, the scattering is evaluated by investigating the kawhere a characterizes the radius of the spherical scatterer and k isthe wavenumber (Table 4.1). As the ultrasound pulse encounters aninterface of two media layers having different acoustical impedancevalues (Z) with one layer being thicker than the wavelength, thenreflection and transmission phenomena take place (Table 4.1). Thegreater the difference in the Z values of the materials, the greateris the reflection coefficient (RC) and the smaller is the transmissioncoefficient (TC). The reflection and transmission coefficients are alsodependent on the angle of incidence according to Snell’s law (Table4.1). As the particle bindings in soft tissues are weak, only longi-tudinal waves can propagate. Thus, the angle of the incident waveis equal to the angle of the reflected wave and the wave proceedingthrough the interface becomes refracted. However, when there isan interface between soft tissue and bone (which has strong parti-cle bindings), also shear waves are encountered (Figure 4.1.).


ii

ii

ii

ii


Incident longitudinal wave Reflected longitudinal wave

Refracted longitudinal wave

2

11

Liquid

Liquid

a)

Incident longitudinal wave Reflected longitudinal wave

Refracted longitudinal wave

3

11

2

Refracted shear wave

Liquid

Solid

b)

Incident longitudinal wave

Leaky surface wave

Liquid

Solid

1 1

Reflected longitudinal waves

c)

Figure 4.1: Snell’s law. a) Refraction and reflection of incident longitudinal wave in aliquid-liquid interface. b) Refraction and reflection of incident longitudinal wave in liquid-solid interface. c) An illustration of leaky surface wave formation.

When the angle of incidence is increased to, or it is greater thanthe critical angle, the wave will not propagate through the interface.Instead, total reflection takes place. Thus, an leaky surface wave isformed that will propagate in parallel to the surface of the interfaceand will reflect longitudinal waves back into the soft-tissue (Figure4.1.). This phenomenon can be measured by an axial transmissiontechnique enabling the evaluation of cortical bone properties in vivo[115, 116].

4.4 ULTRASOUND ABSORPTION

Absorption is dependent on the frequency content of the soundwave and density, viscosity and temperature of the media (Table4.1). The absorbed energy of the acoustic pulse typically is con-


ii

ii

ii

ii


verted into heat. The attenuation coefficient is the sum of absorp-tion, scattering, reflection and beam spreading, which are also closelyrelated to the structure and viscous properties of the media. The ex-ponential decrease in the ultrasound energy along the propagationcan be quantified by the attenuation coefficient (Table 4.1). In acous-tically isotropic materials, the acoustic properties are similar in alldirections. However, biological tissues are commonly anisotropicand this needs to be considered when interpreting ultrasound mea-surements of bone.

4.5 ULTRASOUND APPLICATIONS

Some polar materials, e.g., lead zirconate titanate (PZT), changetheir dimension when a voltage difference is applied on differentsides of the material (electrostriction). Moreover, the voltage differ-ence between the sides of the material can be measured, when thedimensions of the material are changed (piezoelectric effect). Thesematerial properties are utilized in ultrasound transducers when anelectrical pulse is converted to a mechanical pulse (transmission),and when the mechanical pulse is converted back to an electricalsignal (receiving). There are various types of ultrasound transduc-ers which can be matched to the specific applications and needs.

Flat faced ultrasound transducer is technically simple and itis the most widely utilized. The transducer may also be focused,which means that the ultrasound energy is directed to a certain lo-cation through focusing either with on an acoustic lens or by theapplication of phased-array technology (Figure 4.2).


ii

ii

ii

ii


Depth of field

Di FWHM

FL

Lens

RL

Figure 4.2: A schematic view of a focused ultrasound transducer. Focal length (Fl), depthof field (DOF), the diameter of the transducer (Di), radius of lens curvature (Rl), and thebeam width at half maximum amplitude (full width half maximum, FWHM) are indicated.

Table 4.3: Equations describing the characteristics of an focused ultrasound transducer[95, 97].

Parameter Equation

Focal length [m] Fl =Rl

1 − clcm

Full width of the half maximum [m] FWHM =λFlDi

=cmFlf Di

Depth of field [m] DOF = 7λ(FlDi

)2 = 7cm

f(

FlDi

)2

Fl = focal length, Rl = radius of lens curvature, cl = speed of sound in lens,cm = speed of sound in medium, λ = wavelength, Di = diameter of transducerand f = transducer center frequency.

The basic equations describing the focused transducer proper-ties are summarized in Table 4.3. In the pulse-echo (PE) geometry,one transducer is used to transmit and receive the ultrasound sig-nal. All the measurements in this thesis were conducted with PEgeometry. In the through transmission (TT), the geometry trans-mitting transducer is placed on the other side of the object and the


ii

ii

ii

ii


receiving transducer records the transmitted signals on the otherside of the object (Figure 4.3).

Figure 4.3: Transmission and PE geometries. TOF denotes time of flight.

Quantitative ultrasound for evaluation of bone

The majority of the clinical quantitative ultrasound (QUS) devicesthat are intended for use in the diagnostics of osteoporosis mea-sure peripheral sites, e.g., calcaneus, with TT geometry. Although,also radius and phalanges can be measured [117]. Most commonly,the speed of sound and broadband ultrasound attenuation are de-termined [24, 28, 84, 118]. The used techniques to determine speedof sound vary between manufacturers and in an attempt to avoidmanufacturer dependent variation in speed of sound, a form ofstandardization has been proposed [119, 120, 120, 121]. Definitionsdescribing the calculation of various QUS parameters are presentedin Tables 4.4 and 4.6.

Technical and signal processing development in the field of ul-trasound has been adapted also in the QUS evaluation of bone


ii

ii

ii

ii


Table 4.4: Traditional and novel pulse-echo parameters with equations used in the assess-ment bone condition have been collected in the table [27, 122].

Parameter Equation

Integrated reflectioncoefficient [dB]

IRC =1

∆ f∫

∆ f 20 log10Asr( f )Are f ( f )

d f

Apparent integratedbackscatter* [dB]

AIB =1

∆ f∫

∆ f 20 log10Abs( f )Are f ( f )

d f

Broadband ultrasoundbackscatter [dB]

BUB =1

∆ f∫

∆ f 20 log10(Abs( f )Are f ( f )

+ β)d f

Mean of the backscatterdifference** [dB]

MBD =1

∆ f∫

∆ f 20 log10Abs2( f )Abs1( f )

d f

∆ f = analyzed frequency band (typically full width of the half maximum ofthe reference spectrum), Asr = amplitude spectrum of the signal gated at thesurface reflection, Are f = amplitude spectrum of the perfect reflector, Abs =amplitude spectrum of the signal gated at the backscatter and β = attenuationcompensation term. In the MBD the subscripts 1 and 2 refer to delay at thegated backscatter, where 1 is kept constant and 2 is delayed.* Frequency slope of apparent integrated backscatter (FSAB) is determined asa slope of the linear part of the AIB.**Slope of mean of the backscatter difference (SBD) is determined as a slope ofthe linear part of the MBD.

properties. During the last decade, axial transmission techniquesutilizing an array of probes have been applied to measure boneproperties at peripheral sites, e.g., radius and tibia in vivo [29, 115,116, 123–125]. However, the increasing interest of measuring themost serious fracture site, i.e., femoral neck, has stimulated thedevelopment of TT based devices capable of conducting measure-ments at central sites (FemUS) [126–133]. Recently, a photoacous-tic technique has been combined with the axial transmission tech-nique, with the aim of evaluating bone properties, e.g., corticalthickness [134]. This approach has been developed to overcomeproblems in the axial transmission technique originating from the


ii

ii

ii

ii


attenuation and distortion in the ultrasound pulse arising from thesoft tissue overlying the bone [135].

An important development in the single probe systems (PE-mea-surements) has been the advances in the signal processing ca-pabilities. The PE measurement provides reflection and backscat-ter information from the measured object. The technique has beenutilized, e.g., when assessing cortical bone thickness in vivo by mea-suring the reflections from the peri- and endosteal surfaces in longbones in a radial direction [136, 137].

With the PE technique, the integrated reflection coefficient (IRC)and apparent integrated backscatter (AIB) parameters are tradi-tionally measured and their relationship to bone properties canbe investigated [88, 89, 91, 138–145]. More recently, the time slopeof apparent backscatter (TSAB), the frequency slope of apparentbackscatter (FSAB) and the mean of the backscatter difference spec-trum (MBD) parameters were adopted in the evaluation of bonebackscatter [122, 143, 144, 146–148].

4.6 SCANNING ACOUSTIC MICROSCOPY

The elastic properties in biological tissues are anisotropic. By ne-glecting the viscous component, i.e., the time dependent behav-ior, the elastic properties define the anisotropic mechanical prop-erties of the material and can be presented in a six times six matrixcontaining the stiffness tensor. The basic equations for definingthe anisotropic material mechanical behavior are presented in Ta-ble 4.2. The properties of materials are traditionally evaluated byphysical measurements, e.g., three point bending and indentationmethods. Physical measurement enables also the evaluation of vis-coelastic behavior of material. Highly local evaluation of materialproperties, e.g., in bone can be achieved by nano-indentation meth-ods [149–151]. However, the measurement time is long, and thus itis not feasible to obtain coverage of large areas. Recently, a micro-Brillouin scattering technique has been introduced to permit theevaluation of bone elastic properties in thin sections [152, 153].


ii

ii

ii

ii


As described earlier, the ultrasound reflection depends on thematerial density and elastic properties. By measuring the acous-tic impedance and applying information about the material densitythen one can derive the elastic tensor [154]. The advantage of us-ing scanning acoustic microscopy over the mechanical indentationmethods is that one can determine the elastic coefficient of the ma-terial quickly and locally over large areas without damaging thesurface of the sample. Scanning acoustic microscopy (SAM) hasbeen used to estimate calcified tissue elastic properties from the celllevel to the macroscopic level [155]. As an example, a stiffness coef-ficient map in the longitudinal direction (C33) of a human femoralneck cross-section obtained with SAM is presented in Figure 4.4.Moreover, the mechanical properties extracted by SAM have beenapplied in a finite element model [154]. In this thesis, longitudinalelastic coefficients and porosity of human femoral neck and shaftcortical and trabecular bone were measured with SAM. Further-more, the relationships between elastic coefficient, porosity and ca-daver age were evaluated.

500

1000

1500

2000

2500 15

20

25

30

35

40

45

50

55

MPa

c33

Neck

Shaft

Figure 4.4: Cross-sections extracted from the femoral neck and shaft are indicated withdashed lines in the X-ray image (left). Elastic coefficient C33 map obtained with SAM (50MHz, pixel size 16 µm x 16 µm) for a cross-sectional sample of the femoral neck (center)and shaft (right) of a 49 year-old male cadaver.

SAM is based on ultrasound PE measurement with a scanningpositioning stage in the XY-direction. The typical SAM transduc-


ii

ii

ii

ii


ers utilized in bone research work with frequencies approximatelyfrom tens of megahertz to one gigahertz [150, 156, 157]. Moreover,the transducers are heavily focused to obtain a small spot size in or-der to conduct highly local measurements. Thus, in order to reachthe focal region of the transducer, the distance in the Z-directionneeds to be adjustable. The sample surface and the transducerscanning plane must be parallel, which is typically achieved with agoniometer. Depending of the transducer properties, e.g. depth offield, the sample preparation may be challenging. When using highfrequencies and small aperture transducers (over 50 MHz), the bio-logical tissues are usually embedded, e.g., in polymethyl methacry-late (PMMA), and the surfaces of the sample are ground with adecreasing grain size. Furthermore, to ensure the minimal surfaceroughness and maximal signal (without losing energy through scat-tering) from the sample surface, cloths with grinding liquids con-taining diamond particles are used. This type of finishing resultsin mirror-like surfaces (Figure 4.5). The loss of amplitude due to asample’s surface that is not perfectly flat and non-optimal position-ing of the sample (not perfectly parallel surfaces between samplesurface and the transducer scanning plane) may be compensatedfor by using time of flight defocus correction [155, 156].


ii

ii

ii

ii


Figure 4.5: Polymethyl methacrylate (PMMA) embedded bone samples after grinding andpolishing procedure, ready for scanning acoustic microscopy measurements.

4.7 DUAL FREQUENCY ULTRASOUND (DFUS) TECHNIQUE

Bone QUS measurements have conventionally been applied only forproximal sites, e.g., calcaneus, radius or phalanges. However, if onewishes to improve the fracture prediction at the most serious frac-ture sites, e.g., proximal femur or vertebra, one needs to conductsite specific measurements [30–32]. At the central locations, theoverlying soft tissue attenuates and distorts the ultrasound pulse.Variations in soft tissue thickness and composition between mea-surement locations and individuals introduce measurement errorswhich limit the accuracy of bone QUS measurements at the centralsites. To tackle this problem, the dual frequency ultrasound (DFUS)technique was introduced (Figure 4.6) [33, 158].


ii

ii

ii

ii


[V]

Figure 4.6: The analysis window of the ultrasound signal is located at the echo arisingfrom the surface of bone (marked with dashed lines in the figure). The DFUS technique isused to minimize the error arising from the adipose and lean tissues in the determinationof the properties of bone tissue. (Figure is modified from [33]).

With the DFUS technique, one attempts to obtain an estimationof the thickness and composition of the soft tissue that is overlyingthe bone. This information can be extracted at the same time whenconducting the pulse-echo measurement of bone. Furthermore, theobtained information of the soft tissue thickness and compositioncan be used for attenuation compensation of the bone QUS param-eters. This provides a more accurate QUS estimation of the bonestatus at the most important fracture sites. Two assumptions areinherent in the DFUS technique. First, the reflection from the bonesurface is assumed to be frequency independent, and second thesoft tissue is considered to consist of two mediums (adipose andlean tissues). The roughness of the bone - soft tissue interface mayhave a frequency dependent effect on ultrasound reflection. How-ever, in a previous study, the assumption about the frequency inde-pendence of reflection was found not to induce any significant erroron the determined soft tissue composition in vivo [34]. By utilizinga priori information of the attenuation coefficients and speeds ofsound in adipose and lean tissues, the thickness and composition ofthese soft tissues can be estimated by analyzing the measured spec-


ii

ii

ii

ii


trum in two separate frequency bands (Figure 4.7). The derivationof the equations related to DFUS-technique have been presentedin detail elsewhere [33]. The fundamental equations for the DFUStechnique are summarized in Table 4.5.

[V]

Figure 4.7: The spectra of the reference and the sample signals, which are used in theDFUS correction, e.g., of the IRC parameters. Light gray vertical lines denote FWHM ofthe reference and the dark gray vertical lines denote 1 MHz inside the FWHM. Due tonoise, the mean from the area at the 1 MHz frequency bands may provide more accuratevalues for low and high amplitudes.

The DFUS technique has been tested in vitro and in a case studyin vivo [33,34,137]. However, the ultimate performance in determin-ing the soft tissue thickness and composition correctly, and subse-quently in correcting for the error induced by the soft tissue is notknown. Thus, this was evaluated in this thesis with the help ofnumerical models. Focused ultrasound transducers have been ap-plied in order to maximize the signal to noise ratio when measur-ing bones underneath a thick layer of soft tissue, and to localize themeasurement in certain region at the bone surface. With transduc-ers having a fixed focus, the focus may not be optimally located atthe bone surface and this may generate uncertainties when deter-


ii

ii

ii

ii


mining the composition and thickness of the overlying soft tissueusing the DFUS technique. In this thesis, uncertainties in the DFUScalculations related to the non-optimal focusing to the bone sur-face were investigated by conducting experimental measurementsin vitro and numerical simulations.

Table 4.5: Equations related to the dual frequency ultrasound technique. [33, 34]

Parameter Equation

Amplitudelow

Asample,l = Hle−2αadipose,l xadipose e−2αlean,l xlean Are f ,l

Amplitudehigh

Asample,h = Hhe−2αadipose,h xadipose e−2αlean,h xlean Are f ,h

Time offlight [m/s]

∆t = 2(xleanclean

+xadipose

cadipose)

Thicknessadipose [m]

xadipose = (∆t2

+xleanclean

)cadipose

Thicknesslean [m]

xlean =

lnAsample,l

Are f ,l− ln

Asample,h

Are f ,h− cadipose(αadipose,h − αadipose,l)∆t

2(αlean,h − αlean,l)−2cadipose(αadipose,h − αadipose,l)

clean

CorrectionIRC [dB]

IRCcorr = IRCuncorr + 2xadiposeαadipose + 2xleanαlean + ki

x = thickness, H = the term including ultrasound reflections from different surfaces,c = speed of sound, α = attenuation coefficient, ∆t = time of flight back and forthbetween the transducer and the bone surface and ki= correction factor for compensationof ultrasound reflection at the adipose-lean tissue interface. Subscripts adipose andlean refer to adipose and lean tissue properties, respectively. The subscripts l and hrefer to the mean value of the reflection amplitude at 1 MHz-wide frequency bands,which are located at the low and high frequency ends of the FWHM of the referencesignal spectrum.


ii

ii

ii

ii


Table 4.6: Ultrasound through transmission equations have been collected to followingtable [27].

Parameter Equation

Bone SOS [m/s] cb =cwxb

xb − (∆tcw)

Attenuation*[dB/MHz]

BUA =1xb

(20log10Are f ( f )Ab( f )

+ 20log10(TCwb( f )TCbw( f )) )

AA [dB/cm] AA =1

xb∆ f∫

∆ f (20log10Are f ( f )Ab( f )

+ 20log10(TCwb( f )TCbw( f )) )df

c = speed of sound, x = thickness, A( f ) = amplitude spectrum, TC = transmis-sion coefficient and f = frequency, BUA = broadband ultrasound attenuationand AA = average attenuation The subscripts w, b, ref, wb and bw refer to water,bone, reference, water-bone interface and bone-water interface, respectively.*normalized broadband ultrasound attenuation (nBUA [dB/cm/MHz]) is de-termined as a slope of the linear part of the BUA normalized with the bonethickness.

4.8 NUMERICAL MODELLING OF ACOUSTIC WAVEPROPAGATION

Physical phenomenon can be described with partial differential equa-tions. For example, these have been used for modeling of electro-magnetic phenomena and structure’s mechanical behavior. Duringthe last decades the increases achieved in computational power andamount of work memory has made it possible to conduct a numer-ical analysis of the propagating ultrasound wave.

Finite difference method

The finite difference time domain (FDTD) method is a techniquethat can be used to approximate the solutions to partial differentialequations (PDE). The FDTD method has been applied in the eval-uation of ultrasound propagation in bone tissue [159–163]. In theFDTD method, an isotropic element grid is created over the area


ii

ii

ii

ii


of interest. This grid consists of nodes allowing displacement andmovement in relation to their neighbor nodes. The behavior of theproceeding acoustical wave, i.e., the values at the nodes at a certaintime are numerically estimated by approximating the PDE by re-placing the derivative by finite difference approximations. Forward,backward and central difference forms are generally considered inthe FDTD method.

Forward difference ∆ f (x) = f (x + d)− f (x) (4.1)

Backward difference ∇ f (x) = f (x)− f (x − d) (4.2)

Central difference δ f (x) = f (x +12

d)− f (x − 12

d) (4.3)

The derivative of a function y is defined as

y′(x) = limd→0

y(x + d)− y(x)d

(4.4)

where x is the node and d is the difference. With the assumptionthat the difference is finite, the equation can be written as

y′(x) =y(x + d)− y(x)

d=

∆y(x)d

(4.5)

Commercial Wave 2000 plus 3.00 FDTD software was applied inthe present study. Detailed information of the numerical solutioncan be found in Schechter et al. [164]. The two-dimensional acousticwave equation may be derived starting from the hyperbolic partialdifferential equation

∂2w∂t2 = c2∇2w, (4.6)

where c and w represents sound velocity and two-dimensional dis-placement vector, respectively. The Laplace operator ∇2 f is a sec-ond order differential operator, which is defined as the divergenceof the gradient


ii

ii

ii

ii


∇2 = ∇ · ∇ =∂2

∂x2 +∂2

∂y2 (4.7)

The wave equation in the case of homogenous isotropic elasticsolid can be written as follows [165].

ρ∂2w∂2t

= µ∇2w + (λ + µ)∇(∇ · w) (4.8)

Absorption losses are accounted for by deriving the viscoelasticLamé constants λ and µ as [166]

λ → λ +

(ϕ − 2

3η

)∂

∂tand µ → µ + η

∂

∂t(4.9)

Finally, one is are able to write the equation in the form whichthe software uses to solve the two-dimensional acoustic wave equa-tion

ρ∂2w∂2t

=

(µ + η

∂

∂t

)∇2w +

(λ + µ + ϕ

∂

∂t+

η

3∂

∂t

)∇(∇ · w)

(4.10)

where ρ is the density of the material, λ is the first Lamé con-stant, µ is the second Lamé constant, η is the shear viscosity, ϕ isbulk viscosity, t is time, ∇ is the gradient operator, ∇· is the di-vergence operator, and ∂ denotes the partial differential operator.w(x, y, t) is a two-dimensional displacement vector, whose compo-nents x and y are components of displacement of the medium ateach location. The first viscosity (shear viscosity) and the secondviscosity (bulk viscosity) determine the absorption of the soundwave.


ii

ii

ii

ii



ii

ii

ii

ii

5 Aims of the present study

Osteoporosis is a disease that is currently under-diagnosed andunder-treated. This is due to the lack of sensitive and cost-effectivediagnostics tools. Thus, the development of more accurate low-cost portable diagnostic tools, e.g., ultrasound techniques, capableof mass screening of the population for osteoporosis from the mostserious fracture site, e.g., proximal femur would be beneficial. Ul-timately, by identifying patients in need of proper treatment, thenumber of fractures would be reduced leading to great financialsavings to society. If one wishes to achieve this goal, it is necessaryfirst us to understand the error sources and then to evaluate theirsignificance in the ultrasound measurements. Moreover, knowledgeof how bone tissue changes with age and how this is reflected in themeasured ultrasound signal are important.

The specific aims of this thesis were:

• To investigate the relationships between the cortical bone elas-tic coefficients, porosity, tissue type, anatomical location, andthe cadaver age.

• To evaluate the association of traditional and novel ultrasoundbackscatter measurements from intact proximal femurs withthe cadaver age, and the trabecular bone microstructure andbone mineral density.

• To evaluate the performance of the dual frequency ultrasound(DFUS) technique in the determination of the thickness andcomposition of the interfering soft tissue layer, e.g., the layeroverlying proximal femur, and to utilize the information ob-tained in correction of the error arising from the interferinglayer.


ii

ii

ii

ii



ii

ii

ii

ii

6 Materials and methods

The materials and methods applied in studies (I-IV) are summa-rized in this section (Table 6.1).

Table 6.1: Summary of the material and methods utilized in studies I - IV.

Study Samples n Method Parameters

I Human femoral neck 21 in vitro RC → Z → C33cross-sectionsHuman femoral shaft 21 in vitro RC → Z → C33cross-sections

II Intact human 16 ex vivo AIB, FSAB, TSAB, MBDproximal femur

III Cortical bone 4 FDTD model IRCAdipose 1 in silicaLean 1

IV Bovine cortical bone 1 in vitro IRCWater 1Rapeseed oil 1

Cortical bone 1 FDTD model IRCWater 1 in silicaOil 1

IRC = integrated reflection coefficient, RC = reflection coefficient, Z = acouticimpedance, C33 = elastic coefficient in longitudinal direction, AIB = appar-ent integrated backscatter, FSAB = frequency slope of apparent integratedbackscatter, TSAB = time slope of apparent integrated backscatter and MBD= mean of backscatter difference spectrum.

6.1 MATERIALS

In study I, cross-sectional samples from twenty-one male cadav-ers (n = 21, aged 47.1 ± 17.8 years, range 17–82 years) were ob-


ii

ii

ii

ii


tained from Kuopio University Hospital. Ethical approval for col-lection of samples was granted by the National Authority for Medi-colegal Affairs (permission number: 5783/04/044/07). The cadav-ers had no pre-existing conditions that might have affected bonemetabolism. The cross-sectional samples were cut with a bandsaw (KT-210, Koneteollisuus Oy, Helsinki) from the femoral neckand from the shaft underneath the minor trochanter. The sam-ples were dehydrated in ethanol and embedded in polymethyl-methacrylate (PMMA) according to the standard protocols. Thesurfaces of the samples were grounded flat using a grinder (EX-AKT 400CS, ExaktApparatebau, Norderstedt, Germany) with suc-cessively decreasing grain size (ISO/FEPA grit: P500, P800, P1000,P1200 and P4000, Hermes Abrasives Ltd., Virginia Beach, VA andStruers A/S, Ballerup, Denmark), followed by polishing protocolusing cloths moistened with diamond particle suspensions (Phoenix4000, Buehler Ltd., Lake Bluff, IL.). During the measurements, thesamples were immersed in degassed distilled water.

In study II, proximal femurs from sixteen male cadavers wereobtained (n = 16, aged 47.0 ± 16.1 years, range 21 – 77 years). Fourof the samples were obtained from the same cadavers as used instudy I. Thus, the same inclusion criterion and the ethical approvalwas valid for the samples in study II. The samples were freed of softtissue and the DXA and ultrasound measurements were conductedex vivo. Subsequently, the cortical layer was removed using a bandsaw and cylindrical trabecular bone plugs (diameter of 10 mm) wereextracted with a hollow drill bit. The microstructure of the plug wasevaluated using µCT with a 14 µm voxel size (Skyscan-1172, BrukermicroCT, Kontich, Belgium).

In study III, ultrasound propagation was simulated in bone -soft tissue construct with cortical bone densities of 1850, 1900, 1950and 2000 kg/m3 and adipose and lean tissues with densities of 920and 1070 kg/m2, respectively. The material parameters for the sim-ulated tissues were obtained from the literature and the manufac-turer of the simulation software (Wave 2000 plus version 3.00 R3;CyberLogic Inc., New York, NY, USA) [88, 96, 167].


ii

ii

ii

ii

Materials and methods

In study IV, a fresh bovine knee was obtained from a localslaughterhouse (Atria Oyj, Kuopio, Finland). A cortical bone sam-ple was extracted from proximal tibia using a band saw (KT-210,Koneteollisuus Oy, Helsinki) and a low speed diamond blade saw(Buehler Isomet, Buehler Ltd., Lake Bluff, IL). The ultrasound mea-surements of the sample were performed when immersed in de-gassed distilled water, which was overlaid by degassed rapeseedoil (Rainbow, Importer: Inex Partners Oy., Helsinki Finland, manu-factured in Belgium.). The material parameters used in the numer-ical models for bone, water and oil were obtained from the simu-lation software material library. However, the density of the bonewas adjusted to match the density determined experimentally (1962kg/m3).

6.2 METHODS

6.2.1 Basic characteristics of bone

Determination of bone mineral density

The bone mineral density (BMD [g/cm2]) of intact human proximalfemurs was measured with a DXA scanner using a clinical hip mea-surement protocol (Lunar Prodigy, GE Healthcare Ltd., Madison,WI, USA) (Study II). The proximal femurs were positioned accord-ing to the in vivo pelvic anatomy in a plastic container filled withPBS to a depth of 160 mm to mimic the soft tissue. Neck, trochanterand total BMD were evaluated at the standard regions with the soft-ware provided by the manufacturer of the DXA scanner.

Evaluation of trabecular structure

The microstructure of the human trabecular bone samples obtainedfrom femoral neck and shaft was evaluated using a µCT scanner(Skyscan-1172, Bruker microCT, Kontich, Belgium) (Study II). Thevoxel size and imaging parameters applied were 14 x 14 x 14 µm3

isotropic voxel size and tube voltage of 100 kV, tube current of 100µA, 0.5 mm aluminum filter, and 10 repeated scans, respectively.Trabecular bone volume fraction (BV/TV [%]), structural model in-


ii

ii

ii

ii


dex (SMI [-]), average trabecular thickness (Tb.Th [µm]), number(Tb.N [µm−1]) and separation (Tb.Sp [µm]) were calculated. Allanalyses were performed with CTAn software (v. 1.13.2.1, BrukermicroCT, Kontich, Belgium).

Measurement of physical density

The density of the fresh bovine cortical bone was measured with theArchimedes principle, using a precision motion controller to movethe sample (PM500-C Newport, Irvine, CA, USA) and a 1000 g loadcell (Sensotec, Columbus, OH, USA) (Study IV). The measurementwas repeated 10 times. In order to obtain the dry weight, the boneblock was blotted with paper between the measurements.

6.2.2 Ultrasound experiments

The longitudinal elastic coefficients and porosity of the cross-sec-tional bone samples from femoral neck and shaft were investigatedusing scanning acoustic microscopy (SAM) (Study I). The SAMmeasurements were conducted with a system consisting of 200 MHzpulser–receiver (Panametrics 5900PR, Panametrics-NDT, Waltham,MA, USA), a 400 MS/s 12-bit A/D-board (CompuScope 12400,GaGe, Lockport, IL, USA) and a XYZ-scanning stage (Phytron, Willis-ton, VT, USA). The system was controlled with custom-made soft-ware (SAM200Ex, Q-BAM, Halle, Germany). Moreover, a 50 MHzspherical focused transducer (V605/60◦, -6 dB bandwidth 26-64MHz, Valpey Fisher, Hopkinton, USA) with a focal length of 5.2mm and beam diameter at focus of 23 µm was used. Signals wereband-pass filtered using 5 and 90 MHz cutoff frequencies. Thesample was adjusted to be parallel with the transducer face us-ing a custom-made tilting table. Five elastic homogenous mate-rials [Titanium, Suprasil (Heraeus Quarzglas GmbH & Co., KG,Hanau, Germany), polymethylmethacrylate (PMMA), polystyreneand polymethylpentene (TPX)], with known acoustic impedanceswere used as reference phantoms and the time of flight defocuscorrection was applied in all of the measurements. The procedures


ii

ii

ii

ii


related to the SAM measurements have been described in detail inthe literature [155,156,168–170]. The measured acoustic impedancewas converted to an elastic coefficient by applying the followingequation:

Cii = 0.608Z1.923ii (6.1)

where Cii and Zii represent the elastic coefficient vector and acous-tic impedance, respectively [171]. Spatial variations of the elasticand microstructural properties of the cross-sectional samples wereinvestigated. Four regions of interest were defined for both the neck(superior, inferior, anterior and posterior) and the shaft (medial, lat-eral, anterior and posterior). Moreover, the endocortical boundarybetween the cortical and trabecular bone was determined by an ex-perienced histomorphometrist.

Backscatter from intact proximal femurs was measured and thevalues of AIB, FSAB, TSAB and MBD parameters were determined(Study II). The pulse-echo ultrasound measurements were conductedwith an ultrasound system (UltraPAC, Physical Acoustic Co., NJ,USA) consisting of a 500-MHz A/D board, a 0.2 to 100 MHz pulser-receiver board. The ultrasound system was controlled with a cus-tom LabVIEW program (LabVIEW 6.1, National Instrument, Austin,TX, USA). Furthermore, a 5 MHz transducer (V307, PanametricsInc., Waltham, MA, USA) with a 50.9 mm focal distance and 8.3 mmdepth of field (-6 dB) was used. The positioning and the adjustmentof the angle of ultrasound incidence at the bone surface was con-ducted manually. Moreover, to keep the optimal distance betweenthe transducer and the bone surface, a holder with the length of thetransducer focus was attached to the transducer. Intact proximalfemurs immersed in a container filled with a degassed phosphatebuffered saline (PBS) were measured. The backscatter signals wereacquired from the reflection arising from the saline - bone interfacewindowed 0.5 µs after the maximum of the Hilbert-transformed(envelope) signal. The length of each analysis window was set to 1µs (Figure 6.1).

The effect of non-optimal focusing of the ultrasound beam in the


ii

ii

ii

ii


MBD

IRC AIBFSAB

TSAB

Time [µs]

Am

plit

ude [V

]

0 1 2 3 4 5 6 7

0.75

0.5

0.25

0

-0.25

-0.5

-0.75

-1

Figure 6.1: Ultrasound signal measured from femoral neck. The analysis windows andregions for IRC, AIB, MBD and TSAB have been indicated with vertical lines.

determining of the interfering layer thickness and its compositionwith DFUS technique was evaluated (Study IV). Furthermore, theuncertainty arising from non-optimal focusing to the DFUS basedattenuation correction in the measured IRC values was investigated.The measurement system used was the same as in study II. More-over, the same 5 MHz transducer was used. A 5 MHz frequencywas chosen to optimize the spectral attenuation differences and asit is low enough to be used in the future in vivo measurements. Theblock diagram and the measurement setup for study IV are pre-sented in Figures 6.2 and 6.3, respectively. The DFUS analysis hasbeen described detailed in section 4.6.

The surface of the scanned object was adjusted so that it wasparallel to the transducer with a help of a goniometer. The XYZ-scanning device (UK 1I – T24, Physical Acoustics Limited, Cam-bridge, England) enabled adjustment of measurement location andthe distance of the transducer from the sample. The reference reflec-tion measurements were conducted using a polished steel plate im-


ii

ii

ii

ii


Figure 6.2: Block diagram of the DFUS system with an example signal from the experi-mental measurements. The signal represents an ultrasound reflection from cortical bonesurface through 40 mm of rapeseed oil and 10 mm of water.

mersed in degassed distilled water. The distance between the trans-ducer and the steel plate was varied from 30.9 mm to 70.9 mm (± 20mm from the focus of the transducer). Furthermore, the bone sam-ple was measured with overlying interfering medium composedof rapeseed oil and water with a total thickness of 30.9 mm to 70.9mm. Moreover, the composition of the interfering medium was var-ied. The attenuation and speed of sound for the rapeseed oil weredetermined in TT geometry and were applied in the DFUS calcu-lations. The integrated reflection coefficient (IRC) was calculatedfor bone with and without the attenuation compensation using theDFUS obtained information on the interfering layer thickness andcomposition.

6.2.3 Numerical simulations

Wave 2000 plus commercial software (Cybelogic V.3.00; Inc., NewYork, NY, USA) was used in the present FDTD simulations. In stud-


ii

ii

ii

ii


Figure 6.3: The following setup was used in study IV. Bovine cortical bone was immersedin distilled degassed water which was overlaid with a layer of rapeseed oil. A goniometerwas used to keep the sample and transducer surface parallel to each other.

ies III and IV, the transducers were set to have a fixed focus at 30.0mm and 50.9 mm, respectively. In the models, the transducers wereplanar and consisted of a phased array of individual elements. Sim-ulation of a geometrically focused transducer was realized by phas-ing the action of adjacent elements accordingly when transmittingand receiving ultrasound pulses. The simulated attenuation spec-trum was obtained by through transmission geometry for adiposeand lean tissues (Study III) and oil and water (Study IV). In orderto minimize reflections at the outer boundaries of the geometries,infinite boundary conditions were applied. The time step scale wasset to 1.0 in the simulation setup and the resolving wavelength wasadjusted to 149.24 µm, corresponding to a frequency of 9.7 MHz.With these settings, the number of grid elements in the mesh cor-responded to the total number of pixels in the original object ge-ometry file. The grid element-per-cycle ratio was set to 10. The


ii

ii

ii

ii


attenuation spectra were normalized with the material thicknessand the obtained material specific attenuation [dB/cm] was furtherapplied in the DFUS calculations. The TT reference measurementsimulations were performed through a layer of water. Moreover, thePE reference signal was obtained from the simulation in which thedistance from the transducer to water - air interface correspondedthe distance of the transducer to bone surface in the actual simu-lations. The integrated reflection coefficient (IRC), describing thereflection intensity of the ultrasound from the bone surface, wasanalysed from the simulated signals. The IRC values were atten-uation compensated by using information of the interfering layerthickness and the composition extracted with the DFUS technique.

The materials used in the simulations were modeled as homoge-nous and isotropic. The material properties were obtained from theliterature and the simulation software material library (Table 6.2).In order to enhance the stability of the simulation, the simulationparameters were set such that the number of grid elements in themesh corresponded to the total number of pixels in the geometryfile (15 µm pixel size).

In study III, the cortical bone was modeled with four differentdensities (1850, 1900, 1950 and 2000 kg/m3), corresponding to thevalues reported from human cortical bone [42, 43]. The bone wasoverlaid by 30 mm of soft tissue composed of adipose and lean tis-sue layers with various thicknesses. The composition of soft tissuewas varied between 0 and 100 % of adipose tissue. The angle ofultrasound incidence at the interface between the adipose and leantissue layer was set to 0, 5, 10, 15, 20 and 45 degrees. Moreover, theangle of ultrasound incidence at the soft tissue - bone interface wasset from zero to eight degrees with steps of one degree.

In study IV, the density of the modeled cortical bone was setto match that determined experimentally (1962 kg/m3). The non-optimal focusing was modeled such that the distance between thetransducer (having fixed focus of 50.9 mm) and the sample wasvaried from 30.9 mm to 70.9 mm (i.e. -20 mm to +20 mm from thetransducer focal distance) in steps of 2 mm. The reference signal


ii

ii

ii

ii


Table6.2:

Summ

aryof

them

aterialproperties

usedin

thesim

ulationsof

studiesIII

andIV.

Them

aterialproperties

were

obtainedfrom

theliterature

andfrom

thesim

ulationsoftw

arem

ateriallibrary[88,96,167]

Bone

1B

one2

Bone

3B

one4

Adipose

LeanB

one5

Oil

Water

Density

[kg/m

3]1850

19001950

2000920

10701962

9221000

Poissonratio

0.3700.370

0.3700.370

0.4990.499

0.3700.500

0.500

Lateralvelocity[ m

/s]

29003039

31773317

14461596

32141462

1480

Shearvelocity

[m/

s]1300

13801443

150766

681460

00

FirstLam

econstant

[MP

a]9306

1030411561

129221916

271511900

19702190

SecondLam

econstant

[MP

a]3127

36204062

45404

54181

00


ii

ii

ii

ii


was simulated with corresponding distances from the transducerto the water - air interface.

6.2.4 Statistical analysis

Table 6.3: Summary of the statistical methods used in studies I - IV.

Study Statistical test

I Linear regressionOne-Way ANOVA followed by post hocmultiple comparison Tukey–Kramer testsTwo-Way ANOVA

II Linear regressionOne-Way ANOVA followed by post hocmultiple comparison Tukey–Kramer testsCVRMS

III Linear regressionMann-Whitney U-test

IV Linear regression

The statistical analyses in studies I-IV were conducted with acustom-made MATLAB software utilizing the Statistics Toolbox andthe SPSS (SPSS V.14 and 19; Inc., Chicago, IL, USA). The statisticaltest used in the studies are summarized in Table 6.3. Through-out studies I-IV, the associations between different parameter val-ues were evaluated using linear regression. The limit for the sta-tistical significance was set to p < 0.05 for all tests. In study I,the differences in the elastic coefficients between the tissue types(trabecular and cortex), and skeletal sites were evaluated by one-way analysis of variance (ANOVA) followed by post hoc multiplecomparison Tukey–Kramer tests. Moreover, the same statistical testwas used when assessing the significances of difference in corti-cal porosity and elastic coefficient between the quadrants of the


ii

ii

ii

ii


cross-sections of neck and shaft. The interactions of tissue type andskeletal sites on the variation of the elastic coefficient were investi-gated with two-way ANOVA. In study II, the differences betweenthe measured values at different anatomical sites were evaluatedwith one-way analysis of variance (ANOVA) followed by the posthoc Tukey-Kramer multiple comparison tests. The reproducibilityof backscatter measurements was evaluated as the root mean squarecoefficient of variation (CVrms) [172]. In study III, differences be-tween IRC values obtained from bones with various densities weretested with the nonparametric Mann-Whitney U-test.


ii

ii

ii

ii

7 Results

7.1 ELASTIC COEFFICIENT AND POROSITY IN HUMANCORTICAL BONE

The longitudinal elastic coefficient and the porosity of bone wereevaluated by means of SAM in the human femoral neck and shaftin vitro (Study I,). The elastic coefficient of the cortical tissue wassignificantly (F = 112.7, p<0.01) higher than that of trabecular bone.Furthermore, the elastic coefficient was significantly (F = 39.7, p<0.01)higher in the femoral shaft than in the neck (Figure 7.1 a). More-over, the values of elastic coefficient and porosity (Figure 7.1 b) var-ied significantly (p<0.05) within the cross-sections of the neck andshaft.

The elastic coefficients of cortical and trabecular bone were sig-nificantly correlated in the femoral neck (R2 = 0.56, p<0.01) (Figure7.2 a) and shaft (R2 = 0.40, p<0.01). A significant correlation (R2 =0.63, p<0.01) was found between the elastic coefficients of corticaland trabecular bone in the femoral neck and shaft (Figure 7.2 b).


ii

ii

ii

ii


28

30

32

34

36

38

40

42

Neck ShaftTrabecular Cortex Trabecular Cortex

a)

C33

[GP

a]

5

10

15

20

25

30

35

NeckAnterior Posterior Superior Inferior

ShaftAnterior Posterior Medial Lateral

b)

Cort

ex p

oro

sity [

%]

Figure 7.1: Values of elastic coefficient (C33) varied significantly (p<0.05, Tukey Kramer)between skeletal locations and tissue types a). Moreover, cortical porosity varied signif-icantly (p<0.05) between quadrants of cross-sections of femoral neck and shaft b). Eachgroup contains values from all cadavers (n = 21). The bars on top of the figure indicatesignificant (p<0.05) differences between the groups. The gray crosses represent outliers.


ii

ii

ii

ii

Results

Figure 7.2: Elastic coefficients (C33) of cortical and trabecular bone were significantlycorrelated in femoral neck a). Furthermore, elastic coefficients of neck and shaft weresignificantly correlated b). The dots and error bars in a) and b) represent the weightedmeans and standard errors, respectively, of the mean values measured within the fourquadrants of each cross-section.

7.2 ULTRASOUND BACKSCATTER MEASUREMENT OFPROXIMAL FEMUR EX VIVO

The relationships between the backscatter parameters, trabecularbone microstructure, bone mineral density and age were evalu-ated ex vivo (Study II). Ultrasound backscatter parameters measured


ii

ii

ii

ii


from the femoral neck correlated significantly with the values ofneck BMD and total BMD. The highest correlation was observed be-tween the neck MBD and BMD (R2 = 0.45, p<0.01, Figure 7.3) (StudyII). Several ultrasound backscatter parameters correlated with thetrabecular microstructure at the femoral neck, the highest corre-lation being between bone volume fraction (BV/TV) and mean ofbackscatter difference (MBD), R2 = 0.39, p<0.01 (Table 7.1) (StudyII). The reproducibility of the ultrasound backscatter measurementswas calculated using root mean square of coefficient of variation(RMSCV) for ten repeated ultrasound measurement RMSCV = 7.0 -56.3 %.

Table 7.1: Pearson correlation coefficients (R2) between µCT and backscatter parametersat human femoral neck.

Ultrasound BV/TV [%] Tb.Th [µm] Tb.Sp [µm] Tb.N [mm] SMI [-]

AIB [dB] 0.33* ns. 0.43** 0.36* 0.27*FSAB [dB/MHz] 0.37** 0.27* ns. ns. 0.29*TSAB [dB/µs] 0.33* (0.44*) ns. (ns.) ns. (0.33*) ns. (0.41**) ns. (0.45**)MBD [dB] 0.39* ns. 0.33* 0.36* 0.36*

BV/TV = Trabecular bone volume fraction, Tb.Th = average trabecular thick-ness, Tb.N = trabecular number, Tb.Sp = trabecular separation, SMI = struc-tural model index, AIB = apparent integrated backscatter, FSAB = frequencyslope of apparent integrated backscatter, TSAB = time slope of apparent inte-grated backscatter and MBD = mean of backscatter difference.* p < 0.05, ** p < 0.01, and ns.: not significant.The correlation with one outlier in the FSAB parameter is presented in theparenthesis.


ii

ii

ii

ii

Results

Figure 7.3: Mean of backscatter difference (MBD) as a function of bone mineral density(BMD) at the femoral neck. MBD was significantly correlated with the neck BMD.

7.3 AGE-RELATED CHANGES IN BONE

The age-related changes in bone elastic coefficient, porosity, mi-crostructure and ultrasound backscatter were evaluated (Studies Iand II). The elastic coefficients of trabecular and cortical bone inthe femoral neck and shaft were found to increase with age (Figure7.4 a). Furthermore, the porosity of cortical bone in femoral shaftincreased with age (Figure 7.4 a) (Study I).

Most microstructural parameters, determined in the femoral neck,changed with age (range R2 = 0.27–0.48, p<0.05) (Table 7.2) (StudyII). Furthermore, all backscatter parameters determined for the femo-ral neck were significantly (p<0.05) related to the cadaver age (Table7.2) (Study II).


ii

ii

ii

ii


Figure 7.4: Age-related changes in bone. Age accounted for a significant part of thevariation in the values of elastic coefficient (C33) at different anatomical sites (n = 21 foreach group) a). Moreover, cortical porosity in the shaft increased with age b).


ii

ii

ii

ii

Results

Table 7.2: Pearson correlation coefficient (R2) between the cadaver age, µCT and backscat-ter parameters at human femoral neck.

Technique Parameter Age

µCT BV/TV [%] 0.48**Tb.Th [µm] 0.27*Tb.Sp [µm] 0.34*Tb.N [mm] 0.38**SMI [-] ns.

Ultrasound AIB 0.61**FSAB 0.41**TSAB 0.42**MBD 0.41**

BV/TV = Trabecular bone volume fraction, Tb.Th = average trabecular thick-ness, Tb.N = trabecular number, Tb.Sp = trabecular separation, SMI = struc-tural model index, AIB = apparent integrated backscatter, FSAB = frequencyslope of apparent integrated backscatter, TSAB = time slope of apparent inte-grated backscatter and MBD = mean of backscatter difference.* p < 0.05, ** p < 0.01, and ns.: not significant.

7.4 PERFORMANCE OF THE DUAL FREQUENCY ULTRA-SOUND TECHNIQUE

DFUS estimated adipose and lean tissue contents were highly corre-lated with their set values when the tissue interfaces were parallelwith the surface of the transducer, and the bone surface was lo-cated at the transducer focus (R2 = 0.99 and p < 0.001) (Study III).Non-perpendicular ultrasound incidence at soft tissue interfaces in-duced inaccuracy in the evaluation of the soft tissue (Figure 7.5 aand b) (Study III). Regardless of the focusing, the DFUS calculatedtotal thickness of the interfering layer were highly correlated withthe true total thickness of the interfering layer (R2 = 0.99 and p <0.001) (Study IV). However, the accuracy of the estimated compo-sition was dependent on the focusing, both experimentally and insimulations (Figures 7.6 a and b) (Study IV). The thickness valuesof the oil and water layers calculated with the DFUS technique wereover- and underestimated when the transducer was not optimally


ii

ii

ii

ii


focused on the water - bone interface (Study IV). However, the errorwas small when the interface was set optimally at the transducer fo-cus (Figures 7.6 a and b). At the focal distance, the mean errors inthe thickness values of the oil and water layers in the experimentalmeasurements were 1.2 mm (relative error 3.6 %) and 1.5 mm (rel-ative error 12.1 %), respectively (Study IV). In the simulations withthe oil and water, the errors were 0.9 mm (relative error 2.3 %) and0.6 mm (relative error 5.8 %), respectively (Study IV).


ii

ii

ii

ii

Results

0 10 20 30 4010

0

10

20

30

40

50

60

70

Inclination of adipose lean tissue interface [°]

Without reflection frequency dependency correction

With reflection frequency dependency correction

Rela

tive e

rror

in a

dip

ose t

issue t

hic

kness [

%]

a)

0 10 20 30 40

0

5

10

15

20

25

Inclination of adipose lean tissue interface [°]

Without reflection frequency dependency correction

With reflection frequency dependency correction

Rela

tive e

rror

in lean t

issue t

hic

kness [

%]

b)

Figure 7.5: Relative error (mean ± SD) in the thickness of a) adipose- and b) lean tissue asa function of angle of ultrasound incidence at the adipose - lean tissue interface. The datais presented for simulations with four cortical bone densities. By applying a frequencydependent correction of the reflection at the soft tissue - bone interface, the errors werereduced by about 50 %.


ii

ii

ii

ii


30 40 50 60 7040

20

0

20

40

True Interfering Layer Thickness [mm]

Absolute error in oil layer thickness

Absolute error in water layer thickness

0.6 0.8 1 1.2 1.4

Absolu

te E

rror

in D

FU

S C

alc

ula

ted

Inte

rfering L

ayer

Thic

kness [

mm

] Normalized Focal Distance

a)

30 40 50 60 70

10

5

0

5

10

True Interfering Layer Thickness [mm]

Absolute error in oil layer thickness

Absolute error in water layer thickness

0.6 0.8 1 1.2 1.4

Absolu

te E

rror

in D

FU

S C

alc

ula

ted

Inte

rfering L

ayer

Thic

kness [

mm

] Normalized Focal Distance

b)

Figure 7.6: Error in the DFUS-calculated interfering layer thicknesses for experimentalmeasurements a), and simulations b). In both cases, the thickness of the oil and wa-ter layers calculated with the DFUS technique were over- and underestimated when thetransducer was not optimally focused on the surface of the bone, i.e., when the normalizedfocal distance was unity.

The error of IRC arising from attenuation in the interfering layerwas > 300 % (Figure 7.7 a) (Study III). The error decreased to <10 % by applying the attenuation correction based on the DFUSdetermined composition (Study III). However, the inaccuracies inthe determination of soft tissue composition arising from the non-perpendicular angle of ultrasound incidence at interfaces (Figure


ii

ii

ii

ii

Results

7.7 a and b) and non-optimal focusing onto the bone surface (Fig-ure 7.7 c and d) resulted in errors also in the corrected IRC values(Studies III and IV). In the experimental measurements, the errorsin the focal distance were 9.9 dB (no attenuation correction), 1.7 dB(the DFUS correction) and 1.7 dB (correction with the true com-position), respectively. In the simulations, the errors at the focaldistance were 12.5 dB, 0.2 dB and 0.1 dB, respectively (Study IV).

Inclination of soft tissue - bone interface [°]0 2 4 6 8

0

100

200

300

400

500

600IRC

Corrected IRC

Re

lative

err

or

in I

RC

[%

]

a) Inclination of soft tissue - bone interface [°]0 10 20 30 40

0

50

100

150 Without reflection frequency dependency

correction

With reflection frequency dependency

correction

Re

lative

err

or

in I

RC

[%

]

b)

True Interfering Layer Thickness [mm]30 40 50 60 70

-30

-25

-20

-15

-10

-5

0

5

Co

rre

cte

d I

RC

[d

B]

Normalized Focal Distance

c)

Theorethical

DFUS Composition

True Composition

Uncorrected

0.6 0.8 1 1.2 1.4

-30

-25

-20

-15

-10

-5

0

5

True Interfering Layer Thickness [mm]30 40 50 60 70

-30

-25

-20

-15

-10

-5

0

5

Co

rre

cte

d I

RC

[d

B]

Normalized Focal Distance

d)

Theorethical

DFUS Composition

True Composition

Uncorrected

0.6 0.8 1 1.2 1.4

-30

-25

-20

-15

-10

-5

0

5

Figure 7.7: Relative error (mean ± SD) in the corrected integrated reflection coefficient(IRC) values originating from inclination between soft tissue - bone interface a) and adi-pose - lean tissue interfaces b) (data represents simulations with four cortical bone densi-ties). After correcting the frequency dependence of the reflection at the soft tissue - boneinterface, errors were reduced by about 50 %. The effect of non-optimal focusing in thecorrection of IRC using the DFUS-calculated composition (green lines) and the true com-position (blue lines) for in vitro measurements c) and simulations d).


ii

ii

ii

ii



ii

ii

ii

ii

8 Discussion

If one wishes to estimate the fracture risk in osteoporotic patientsreliably, then it is necessary to conduct site specific measurements[30–32]. The anatomical site where the most serious fractures occuris the proximal femur. Thus, the femoral neck is the primary targetfor undertaking ultrasound measurements. Unfortunately, at cen-tral sites, e.g., proximal femur, presence of the overlying soft tissueinduces a significant error in quantitative ultrasound (QUS) mea-surements and this reduces the reliability of the measurements ofcortical and trabecular bone [167, 173–175]. In an attempt o com-pensate for the errors originating from the soft tissue in QUS mea-surements, the dual frequency ultrasound technique (DFUS) wasintroduced [33]. However, the non-perpendicular incidence of ul-trasound pulse at soft tissue and soft tissue - bone interfaces andnon-optimal focusing onto the soft tissue - bone interface can causesignificant errors in the DFUS calculations. Furthermore, it is veryimportant to identify the location within the proximal femur that isoptimal for quantitative ultrasound measurement. Ideally at sucha site, ultrasound measurements are possible and the ultrasoundparameters are sensitive to bone strength, structure and density.Moreover, it is essential to know the age-related changes in bonemicrostructure, mineral density and tissue elastic properties, whichare related closely to the measured QUS parameters as well asthe whole bone mechanical properties at the most serious fracturesites [90, 140]. This is important for understanding the normal age-dependent variation in comparison with the osteoporosis relateddecline in the bone properties. This thesis focused on the abovementioned issues through numerical simulations and experimentalstudies.


ii

ii

ii

ii


8.1 ELASTIC COEFFICIENTS AND POROSITY OF BONETISSUE VARY DURING AGING

Age related changes in the elastic coefficient in the longitudinaldirection and the porosity of cortical and trabecular bone in hu-man proximal femoral neck and shaft were investigated with SAM(Study I). In the shaft, the cortical porosity was found to increasewith age, which is in agreement with the literature [5, 7, 176–179].Moreover, the cortical thickness was found to decline with age,which has also been proposed in earlier publications [4, 8]. To-gether these changes decease the areal bone mineral density, whichis generally used in the diagnostics of osteoporosis. Furthermore,these factors are known to reduce mechanical strength of bone[53, 54, 180]. Also in line with the literature, the cortical porositywas found to be highest in the posterior quadrant of femoral shaftand to manifest the clearest dependence on age [179, 181].

The elastic coefficients of trabecular and cortical bone, and femo-ral neck and shaft were correlated. This indicates that general load-ing of the femur may have a similar effect on the tissue stiffness ofboth the cortical and the trabecular bone matrix, and in both loca-tions investigated (the neck and shaft). Although the values of theelastic coefficient were correlated they were higher in the femoralshaft than in the femoral neck, possibly due to site specific differ-ences in loading conditions [182]. The elastic coefficient was higherin the cortex than in the trabecular bone in the femoral neck andshaft. This may be due to the fact that cortical tissue is directlyexposed to high stresses, whereas the trabecular network mainlysupports the cortical wall preventing the deformation of the cor-tex [77]. These findings are in line with the report of Turner et al.,that the values of elastic modulus of trabecular and cortical bone inthe longitudinal direction are different (EL) [183].

The elastic coefficients of cortical and trabecular bone in thefemoral neck and shaft were found to increase with age. The in-crease in the elastic coefficient is known to be related to age de-pendent changes in bone mineral and organic phases and when an


ii

ii

ii

ii

Discussion

individual ages, the remodeling is disturbed leading to changesin osteoclast and osteoblast activity affecting tissue average age[46, 49, 54, 56]. This may be one of factors explaining why the bonetissue of elderly persons becomes brittle [58].

The human samples investigated in this thesis were extractedfrom male cadavers. Thus, the present results may not be gener-alized to the bones of female individuals as hormonal differencesbetween sex plays a role in bone tissue development and degener-ation. For example, the decline in the female estrogen levels afterthe menopause is known to accelerate the degeneration of bonethus increasing the fracture risk for postmenopausal females [184].

8.2 ULTRASOUND BACKSCATTER IN PROXIMAL FEMUR ISRELATED TO BONE DENSITY AND MICROSTRUCTURE

For the first time ultrasound backscatter parameters were deter-mined ex vivo in intact human proximal femoral neck and trochanter(Study II ). Novel (TSAB and MBD) and traditional (AIB and FSAB)ultrasound backscatter parameters were extracted, and comparedwith areal BMD, trabecular microstructure, and cadaver age. Thebackscatter parameters measured ex vivo at the most important frac-ture sites at the proximal femur correlated with areal BMD. This isencouraging since BMD is typically used when diagnosing osteo-porosis.

Ultrasound backscatter from trabecular tissue was found to de-crease with cadaver age. This may be related to changes in bonemicrostructure with age. These changes include thinning of thecortex and degeneration of the trabecular structure [4, 5, 8, 73, 185].This is supported by reports that the trabecular structure controlsultrasound backscatter [86, 138, 140, 142, 186–188].

There were significant correlations detected between the backscat-ter parameters and trabecular microstructure. The values of backscat-ter parameters obtained from the femoral neck were found to bemore strongly correlated with BMD and trabecular structure thanthe values obtained for the trochanter. If one wishes to estimate


ii

ii

ii

ii


the fracture risk reliably, site specific measurements are warranted[30–32]. Thus, the femoral neck is the primary target to conduct theultrasound measurements. Earlier in vitro studies have describedhigher correlations between ultrasound backscatter and trabecularstructure and BMD than in study II [86, 91, 138, 140–143, 186, 187,189, 190]. In contrast to these reports, the ultrasound backscattermeasurements in study II were performed in situ. Thus, the cor-tical bone distorting and attenuating ultrasound pulse affected therecorded backscatter from trabecular bone. The cortical bone isknown to cause inaccuracy to AIB and FSAB backscatter param-eters and may have influenced the correlations reported in studyII [191]. TSAB and MDB parameters are based on the measure-ment of decay of ultrasound backscatter within bone and thus havebeen claimed to be sensitive to the trabecular bone independentlyof the overlying cortical bone [122]. The novel backscatter param-eter MBD performed equally well as the traditional AIB parameterto estimate the BMD, and slightly better when estimating the tra-becular microstructure. In study II, the ultrasound measurementlocation and the analyzed region of the trabecular microstructuremay not have been perfectly matched as the trabecular bone plugswere extracted for µCT from intact femurs after the ultrasound mea-surements. This may have contributed negatively to the reportedcorrelations.

Further improvements in the measurement technique should beimplemented in order to increase the reproducibility of the backscat-ter measurements. For example, by applying a phased array sys-tem, the perpendicularity of ultrasound pulse incidence at the bonesurface could be optimized. Furthermore, through B-mode imag-ing, the exact measurement location at the bone surface (e.g. cen-ter of the femoral neck) could be verified and the focusing of theultrasound beam could be optimally set to the soft tissue - boneinterface, or even to focus inside the bone in order to maximize thebackscatter from the trabecular structure.


ii

ii

ii

ii

Discussion

8.3 EVALUATION OF THE DUAL FREQUENCY ULTRASOUNDTECHNIQUE

The ability of the DFUS technique to compensate and determinethe interfering layer thickness and its composition was evaluatedwith numerical simulations (Study III). It was found that the totalthickness and composition of the soft tissue covering bone could bedetermined accurately when the ultrasound incidence was perpen-dicular to the tissue interfaces and the soft tissue - bone interfacewas located at the focal distance of the transducer.

Furthermore, the capability of the DFUS technique in compen-sating for the error in IRC, originating from the interfering soft tis-sue layer was investigated numerically. Since the total thickness andcomposition of the soft tissue covering the bone could be accuratelydetermined, also the IRC values of the bone measured through theinterfering soft tissue layer could be accurately corrected. More-over, the DFUS corrected IRC values measured for four sampleswith different densities were shown to be statistically different (p <0.05). The present findings are in line with earlier studies in whichthe DFUS technique has been found to detect accurately the de-crease in adipose tissue percent during dieting and to significantlydecrease the soft tissue induced error in IRC [33, 34].

In vivo the ultrasound incidence at the interfaces between softtissues and soft tissue and bone may not be perpendicular. It wasfound that with increasing deviation from perpendicular incidence,the error in the DFUS determined composition and correction ofIRC values increased. This is logical as the inclined interfaces maycause phase interference affecting the frequency content of the pulseleading to an error in the DFUS based correction of IRC [192]. More-over, it was found that non-perpendicular incidence at the soft tis-sue - bone interface is a more significant source of error than thatat the interface between the soft tissues. This is probably due to thehigher difference in the acoustic impedance at the soft tissue - boneinterface. In order to overcome the problem related to measurementwith a single-element transducer, signals may be recorded from var-


ii

ii

ii

ii


ious angles with only those signals having the highest amplitudes(i.e. perpendicular incidence at the interface) being used for ana-lyzes. Furthermore, when applying a phased array transducer, theinclination can be monitored and minimized based on the B-modeimage.

Numerous sources of inaccuracy are present in vivo [137, 193,194]. For example, the tissue surfaces and interfaces are not per-fectly flat, and the soft tissue is composed of more than two acous-tically different tissue types, e.g., adipose- and lean tissue, skin,tendons and fascia. Moreover, the thickness of the soft tissue variesbetween patients, potentially causing non-optimal focusing of theultrasound beam to the soft tissue - bone interface. The errors inDFUS technique arising from non-optimal focusing of the ultra-sound beam on the soft tissue - bone interface were investigatedwith numerical simulations as well as with in vitro measurements(Study IV). When the soft tissue - bone interface was located atthe focal distance, the errors induced by the soft tissue were min-imized, was as also evident in study III. Furthermore, despite es-timating the total thickness accurately, when the transducer wasnot optimally focused onto the bone surface, the estimation of theinterfering layer composition was inaccurate. This clearly affectedthe accuracy of the correction of the IRC values. To summarize,non-optimal focusing introduces inaccuracies in the determinationof soft tissue composition, which further leads to uncertainty inthe corrected IRC values. However, although the correction of IRCvalues utilizing the DFUS technique is not always perfect, the cor-rected values are much closer to the true IRC values of bone. If onewishes to improve the accuracy of the DFUS technique in vivo, it isrecommended that adjustable focusing should be conducted withthe phased array technique.


ii

ii

ii

ii

9 Summary and conclusions

In this thesis pulse-echo ultrasound methods and their potentialfor evaluating bone properties were investigated. Trabecular andcortical bone elastic coefficients and their variations with age inthe femoral neck and shaft were determined. Furthermore, agedependent changes in the bone mineral density and trabecular mi-crostructure were evaluated with novel and traditional ultrasoundbackscatter parameters from intact proximal femurs ex vivo. More-over, the performance of the dual frequency ultrasound (DFUS)technique in the reduction of the error in the bone QUS measure-ments was investigated. Furthermore, the error sources occurringin ultrasound measurements and their significance with respect todual frequency ultrasound technique were evaluated both numeri-cally and experimentally.

The main conclusions may be summarized as follows:

• The elastic coefficient of calcified matrix in cortical and tra-becular bone in human femoral neck and shaft are different.Moreover, the elastic coefficient and porosity vary between theinvestigated anatomical locations as well as with the anatom-ical planes. The elastic coefficient and cortical porosity bothincrease with age.

• Novel and traditional backscatter parameters can be linkedto the trabecular bone microstructure and the bone mineraldensity in intact proximal femurs ex vivo. Furthermore, thebackscatter parameters are age dependent.

• Non-optimal focusing of ultrasound to soft tissue - bone in-terface and non-perpendicular ultrasound incidence at softtissue and soft tissue - bone interfaces induce a significant er-ror when estimating the soft tissue thickness and compositionwith the DFUS technique.


ii

ii

ii

ii


• The correct application of the DFUS method can significantlyreduce the error induced by attenuation in the interferinglayer overlying the bone.


ii

ii

ii

ii

Summary and conclusions


ii

ii

ii

ii



ii

ii

ii

ii

Bibliography

[1] I. M. van der Sluis, “Reference data for bone density and body composition measuredwith dual energy x ray absorptiometry in white children and young adults,” Archivesof Disease in Childhood 87, 341–347 (2002).

[2] J.-P. Bonjour, T. Chevalley, S. Ferrari, and R. Rizzoli, “The importance and relevance ofpeak bone mass in the prevalence of osteoporosis,” Salud pública de México 51 Suppl1, S5–17 (2009).

[3] H. Haapasalo, P. Kannus, H. Sievänen, M. Pasanen, K. Uusi-Rasi, A. Heinonen, P. Oja,and I. Vuori, “Development of mass, density, and estimated mechanical characteristicsof bones in Caucasian females,” Journal of bone and mineral research : the official journalof the American Society for Bone and Mineral Research 11, 1751–60 (1996).

[4] A. M. Parfitt, “Age-related structural changes in trabecular and cortical bone: cellularmechanisms and biomechanical consequences,” Calcified tissue international 36 Suppl1, S123–8 (1984).

[5] S. A. Feik, C. D. L. Thomas, and J. G. Clement, “Age-related changes in corticalporosity of the midshaft of the human femur,” Journal of anatomy 191 ( Pt 3, 407–16(1997).

[6] W.-Q. Cui, Y.-Y. Won, M.-H. Baek, D.-H. Lee, Y.-S. Chung, J.-H. Hur, and Y.-Z. Ma,“Age-and region-dependent changes in three-dimensional microstructural propertiesof proximal femoral trabeculae,” Osteoporosis international : a journal established as resultof cooperation between the European Foundation for Osteoporosis and the National Osteoporo-sis Foundation of the USA 19, 1579–87 (2008).

[7] D. M. L. Cooper, C. D. L. Thomas, J. G. Clement, A. L. Turinsky, C. W. Sensen, andB. Hallgrímsson, “Age-dependent change in the 3D structure of cortical porosity atthe human femoral midshaft,” Bone 40, 957–65 (2007).

[8] H. Chen, X. Zhou, S. Shoumura, S. Emura, and Y. Bunai, “Age- and gender-dependentchanges in three-dimensional microstructure of cortical and trabecular bone at thehuman femoral neck,” Osteoporos Int 21, 627–636 (2010).

[9] “NIH Consensus Statement - Osteoporosis Prevention, Diagnosis, and Therapy,”(2000).

[10] K. E. Ensrud, “Epidemiology of fracture risk with advancing age,” The journals ofgerontology. Series A, Biological sciences and medical sciences 68, 1236–42 (2013).

[11] S. R. Cummings and L. J. Melton, “Epidemiology and outcomes of osteoporotic frac-tures,” Lancet 359, 1761–7 (2002).

[12] G. S. Keene, M. J. Parker, and G. A. Pryor, “Mortality and morbidity after hip frac-tures,” BMJ (Clinical research ed.) 307, 1248–50 (1993).


ii

ii

ii

ii


[13] E. Hernlund, A. Svedbom, M. Ivergå rd, J. Compston, C. Cooper, J. Stenmark, E. V.McCloskey, B. Jönsson, and J. a. Kanis, “Osteoporosis in the European Union: medicalmanagement, epidemiology and economic burden. A report prepared in collaborationwith the International Osteoporosis Foundation (IOF) and the European Federation ofPharmaceutical Industry Associations (EFPIA),” Archives of osteoporosis 8, 136 (2013).

[14] C. Cooper, “Epidemiology of osteoporosis,” Osteoporosis international : a journal es-tablished as result of cooperation between the European Foundation for Osteoporosis and theNational Osteoporosis Foundation of the USA 9 Suppl 2, S2–8 (1999).

[15] M. J. Barger-Lux and R. R. Recker, “Bone microstructure in osteoporosis: transilialbiopsy and histomorphometry,” Top Magn Reson Imaging 13, 297–305 (2002).

[16] J. A. Kanis, “Diagnosis of osteoporosis and assessment of fracture risk,” Lancet 359,1929–1936 (2002).

[17] J. D. Currey, “Bone architecture and fracture,” Current osteoporosis reports 3, 52–6 (2005).

[18] R. P. Heaney, “Is the paradigm shifting?,” Bone 33, 457–65 (2003).

[19] J. S. Gregory and R. M. Aspden, “Femoral geometry as a risk factor for osteoporotichip fracture in men and women,” Medical engineering & physics 30, 1275–86 (2008).

[20] T. Rikkonen, K. Salovaara, J. Sirola, M. Kärkkäinen, M. Tuppurainen, J. S. Jurvelin,R. Honkanen, E. Alhava, and H. Kröger, “Physical activity slows femoral bone lossbut promotes wrist fractures in postmenopausal women: a 15-year follow-up of theOSTPRE study,” Journal of bone and mineral research : the official journal of the AmericanSociety for Bone and Mineral Research 25, 2332–40 (2010).

[21] S. R. Lord, S. T. Smith, and J. C. Menant, “Vision and falls in older people: risk factorsand intervention strategies,” Clinics in geriatric medicine 26, 569–81 (2010).

[22] “National Osteoporosis Guideline Group - Osteoporosis: Clinical guideline for pre-vention and treatment Executive,” (2014).

[23] T. Nguyen, J. Center, and J. Eisman, “Osteoporosis: underrated, underdiagnosed andundertreated,” Medical journal of Australia 180 (2004).

[24] C. F. Njeh, C. M. Boivin, and C. M. Langton, “The role of ultrasound in the assessmentof osteoporosis: a review,” Osteoporosis international : a journal established as result ofcooperation between the European Foundation for Osteoporosis and the National OsteoporosisFoundation of the USA 7, 7–22 (1997).

[25] D. C. Bauer, C. C. Glüer, J. A. Cauley, T. M. Vogt, K. E. Ensrud, H. K. Genant, and D. M.Black, “Broadband ultrasound attenuation predicts fractures strongly and indepen-dently of densitometry in older women. A prospective study. Study of OsteoporoticFractures Research Group,” Archives of internal medicine 157, 629–34 (1997).

[26] C. C. Glüer, “Quantitative ultrasound techniques for the assessment of osteoporosis:expert agreement on current status. The International Quantitative Ultrasound Con-sensus Group,” Journal of bone and mineral research : the official journal of the AmericanSociety for Bone and Mineral Research 12, 1280–8 (1997).

[27] P. Laugier and G. Haïat, Bone Quantitative Ultrasound (Springer Netherlands, Dor-drecht, 2011).


ii

ii

ii

ii

Bibliography

[28] C. M. Langton, “The 25th anniversary of BUA for the assessment of osteoporosis:time for a new paradigm?,” Proceedings of the Institution of Mechanical Engineers. PartH, Journal of engineering in medicine 225, 113–25 (2011).

[29] M. Määttä, P. Moilanen, P. Nicholson, S. Cheng, J. Timonen, and T. Jämsä, “Correlationof tibial low-frequency ultrasound velocity with femoral radiographic measurementsand BMD in elderly women,” Ultrasound in medicine & biology 35, 903–11 (2009).

[30] G. M. Blake and I. Fogelman, “Fracture prediction by bone density measurements atsites other than the fracture site: the contribution of BMD correlation,” Calcif TissueInt 76, 249–255 (2005).

[31] D. Marshall, O. Johnell, and H. Wedel, “Meta-analysis of how well measures of bonemineral density predict occurrence of osteoporotic fractures,” BMJ (Clinical researched.) 312, 1254–9 (1996).

[32] S. R. Cummings, D. M. Black, M. C. Nevitt, W. Browner, J. Cauley, K. Ensrud, H. K.Genant, L. Palermo, J. Scott, and T. M. Vogt, “Bone density at various sites for pre-diction of hip fractures. The Study of Osteoporotic Fractures Research Group,” Lancet341, 72–5 (1993).

[33] O. Riekkinen, M. A. Hakulinen, J. Töyräs, and J. S. Jurvelin, “Dual-frequencyultrasound–new pulse-echo technique for bone densitometry,” Ultrasound in medicine& biology 34, 1703–8 (2008).

[34] J. P. Karjalainen, J. Töyräs, T. Rikkonen, J. S. Jurvelin, and O. Riekkinen, “Dual-frequency ultrasound technique minimizes errors induced by soft tissue in ultrasoundbone densitometry,” Acta radiologica (Stockholm, Sweden : 1987) 49, 1038–41 (2008).

[35] J. A. Buckwalter, M. J. Glimcher, R. R. Cooper, and R. Recker, “Bone biology. I: Struc-ture, blood supply, cells, matrix, and mineralization,” Instructional course lectures 45,371–86 (1996).

[36] H. Gray, Gray’s Anatomy, 35th ed. (Longman Publishing Group, 1973).

[37] W. Bloom and D. W. Fawcett, A textbook of histology, 9 th ed. (W B Saunders Co, 1968).

[38] L. J. Melton, E. A. Chrischilles, C. Cooper, A. W. Lane, and B. L. Riggs, “Perspective.How many women have osteoporosis?,” Journal of bone and mineral research : the officialjournal of the American Society for Bone and Mineral Research 7, 1005–10 (1992).

[39] E. Seeman and P. D. Delmas, “Bone quality–the material and structural basis of bonestrength and fragility,” The New England journal of medicine 354, 2250–61 (2006).

[40] D. Chappard, M. F. Baslé, E. Legrand, and M. Audran, “New laboratory tools in theassessment of bone quality,” Osteoporosis international : a journal established as result ofcooperation between the European Foundation for Osteoporosis and the National OsteoporosisFoundation of the USA 22, 2225–40 (2011).

[41] J. Wolff, Das Gesetz der Transformation der Knochen (August Hirschwald, Berlin, 1892).

[42] P. J. Atkinson and J. A. Weatherell, “Variation in the density of the femoral diaphysiswith age,” The Journal of bone and joint surgery. British volume 49, 781–8 (1967).

[43] J. Wall, S. Chatterji, and J. Jeffery, “Human femoral cortical bone: a preliminary reporton the relationship between strength and density,” Medical and biological engineering 10,673–676 (1972).


ii

ii

ii

ii


[44] E. Grodum, J. Gram, K. Brixen, and J. Bollerslev, “Autosomal dominant osteopetrosis:bone mineral measurements of the entire skeleton of adults in two different subtypes,”Bone 16, 431–4 (1995).

[45] J. Compston, “Bone quality: what is it and how is it measured?,” Arquivos brasileirosde endocrinologia e metabologia 50, 579–85 (2006).

[46] X. Feng and J. M. McDonald, “Disorders of bone remodeling,” Annual review of pathol-ogy 6, 121–45 (2011).

[47] D. Burr, “The contribution of the organic matrix to bone’s material properties,” Bone31, 8–11 (2002).

[48] U.S. Department of Health and Human Services, “Bone Health and Osteoporosis: AReport of the Surgeon General,” US Dep. Health Hum. Serv (2004).

[49] J. A. Buckwalter, M. J. Glimcher, R. R. Cooper, and R. Recker, “Bone biology. II:Formation, form, modeling, remodeling, and regulation of cell function,” Instructionalcourse lectures 45, 387–99 (1996).

[50] B. F. Boyce and L. Xing, “Biology of RANK, RANKL, and osteoprotegerin,” Arthritisresearch & therapy 9 Suppl 1, S1 (2007).

[51] B. R. McCreadie, S. J. Hollister, M. B. Schaffler, and S. a. Goldstein, “Osteocyte la-cuna size and shape in women with and without osteoporotic fracture,” Journal ofbiomechanics 37, 563–72 (2004).

[52] A. M. Parfitt, “The bone remodeling compartment: a circulatory function for bonelining cells,” Journal of bone and mineral research : the official journal of the AmericanSociety for Bone and Mineral Research 16, 1583–5 (2001).

[53] P. Zioupos and J. D. Currey, “Changes in the stiffness, strength, and toughness ofhuman cortical bone with age,” Bone 22, 57–66 (1998).

[54] A. J. Bailey, T. J. Sims, E. N. Ebbesen, J. P. Mansell, J. S. Thomsen, and L. Mosek-ilde, “Age-related changes in the biochemical properties of human cancellous bonecollagen: relationship to bone strength,” Calcified tissue international 65, 203–10 (1999).

[55] P. Milovanovic, J. Potocnik, D. Djonic, S. Nikolic, V. Zivkovic, M. Djuric, and Z. Rako-cevic, “Age-related deterioration in trabecular bone mechanical properties at materiallevel: nanoindentation study of the femoral neck in women by using AFM,” Experi-mental gerontology 47, 154–9 (2012).

[56] X. Wang, X. Shen, X. Li, and C. M. Agrawal, “Age-related changes in the collagennetwork and toughness of bone,” Bone 31, 1–7 (2002).

[57] A. H. Burstein, D. T. Reilly, and M. Martens, “Aging of Bone Tissue: MechanicalProperties,” The journal of bone and joint surgery 58, 82–6 (1976).

[58] E. A. Zimmermann, E. Schaible, H. Bale, H. D. Barth, S. Y. Tang, P. Reichert, B. Busse,T. Alliston, J. W. Ager, and R. O. Ritchie, “Age-related changes in the plasticity andtoughness of human cortical bone at multiple length scales,” Proceedings of the NationalAcademy of Sciences of the United States of America 108, 14416–21 (2011).

[59] J. Compston, “Clinical and therapeutic aspects of osteoporosis,” European journal ofradiology 71, 388–91 (2009).


ii

ii

ii

ii

Bibliography

[60] S. R. Cummings, P. M. Cawthon, K. E. Ensrud, J. A. Cauley, H. A. Fink, and E. S.Orwoll, “BMD and risk of hip and nonvertebral fractures in older men: a prospectivestudy and comparison with older women,” Journal of bone and mineral research : theofficial journal of the American Society for Bone and Mineral Research 21, 1550–6 (2006).

[61] WHO - Prevention and management of osteoporosis, vol. 921 ed. (WHO Technical ReportSeries, Geneva, 2003).

[62] A. C. Looker, L. G. Borrud, J. P. Hughes, B. Fan, J. a. Shepherd, and L. J. Melton,Lumbar spine and proximal femur bone mineral density, bone mineral content, and bone area:United States, 2005-2008 2012).

[63] WHO - Assessment of fracture risk and its application to screening for postmenopausal osteo-porosis, vol. 843 ed. (WHO Technical Report Series, Geneva, 1994).

[64] J. A. Kanis, L. J. Melton, C. Christiansen, C. C. Johnston, and N. Khaltaev, “TheDiagnosis of Osteoporosis,” Journal of bone and mineral research 9, 1137–1141 (1994).

[65] J. A. Kanis, E. V. McCloskey, H. Johansson, A. Oden, L. J. Melton, and N. Khaltaev,“A reference standard for the description of osteoporosis,” Bone 42, 467–75 (2008).

[66] J. Compston, “Guidelines for the management of osteoporosis: the present and thefuture,” Osteoporosis international : a journal established as result of cooperation between theEuropean Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA16, 1173–6 (2005).

[67] “WHO - Scientific group on the assessment of osteoporosis at primary health carelevel,” (2004).

[68] G. M. Blake, E. J. Harrison, and J. E. Adams, “Dual X-ray absorptiometry: cross-calibration of a new fan-beam system,” Calcified tissue international 75, 7–14 (2004).

[69] E. W. Yu, B. J. Thomas, J. K. Brown, and J. S. Finkelstein, “Simulated increases in bodyfat and errors in bone mineral density measurements by DXA and QCT,” Journal ofbone and mineral research : the official journal of the American Society for Bone and MineralResearch 27, 119–24 (2012).

[70] H. H. Bolotin, H. Sievänen, J. L. Grashuis, J. W. Kuiper, and T. L. Järvinen, “Inac-curacies inherent in patient-specific dual-energy X-ray absorptiometry bone mineraldensity measurements: comprehensive phantom-based evaluation,” Journal of bone andmineral research : the official journal of the American Society for Bone and Mineral Research16, 417–26 (2001).

[71] H. Hull, Q. He, J. Thornton, F. Javed, L. Allen, J. Wang, R. N. Pierson, and D. Gal-lagher, “iDXA, Prodigy, and DPXL dual-energy X-ray absorptiometry whole-bodyscans: a cross-calibration study,” Journal of clinical densitometry : the official journal ofthe International Society for Clinical Densitometry 12, 95–102 (2010).

[72] J. E. Adams, “Quantitative computed tomography,” European journal of radiology 71,415–24 (2009).

[73] H. M. Macdonald, K. K. Nishiyama, J. Kang, D. a. Hanley, and S. K. Boyd, “Age-related patterns of trabecular and cortical bone loss differ between sexes and skeletalsites: a population-based HR-pQCT study,” Journal of bone and mineral research : theofficial journal of the American Society for Bone and Mineral Research 26, 50–62 (2011).


ii

ii

ii

ii


[74] H. T. Kokkonen, J.-S. Suomalainen, A. Joukainen, H. Kröger, J. Sirola, J. S. Jurvelin,J. Salo, and J. Töyräs, “In vivo diagnostics of human knee cartilage lesions usingdelayed CBCT arthrography,” Journal of orthopaedic research : official publication of theOrthopaedic Research Society 1–10 (2013).

[75] H. Isaksson, J. Töyräs, M. A. Hakulinen, A. S. Aula, I. Tamminen, P. Julkunen,H. Kröger, and J. S. Jurvelin, “Structural parameters of normal and osteoporotichuman trabecular bone are affected differently by microCT image resolution,” Os-teoporosis international : a journal established as result of cooperation between the EuropeanFoundation for Osteoporosis and the National Osteoporosis Foundation of the USA 22, 167–77(2011).

[76] K. E. S. Poole, G. M. Treece, P. M. Mayhew, J. Vaculík, P. Dungl, M. Horák, J. J. Štepán,and A. H. Gee, “Cortical thickness mapping to identify focal osteoporosis in patientswith hip fracture,” PloS one 7, e38466 (2012).

[77] C. D. L. Thomas and P. Mayhew, “Femoral neck trabecular bone: loss with aging androle in preventing fracture,” Journal of Bone and . . . 24, 1808–1818 (2009).

[78] K. Engelke, C. Libanati, T. Fuerst, P. Zysset, and H. K. Genant, “Advanced CT basedin vivo methods for the assessment of bone density, structure, and strength,” Currentosteoporosis reports 11, 246–55 (2013).

[79] M. L. Bouxsein and E. Seeman, “Quantifying the material and structural determinantsof bone strength,” Best practice & research. Clinical rheumatology 23, 741–53 (2009).

[80] S. P. Väänänen, J. S. Jurvelin, and H. Isaksson, “Estimation of 3-D shape, internaldensity and mechanics of proximal femur by combining bone mineral density imageswith shape and density templates,” Biomechanics and modeling in mechanobiology 11,791–800 (2012).

[81] J. E. M. Koivumäki, J. Thevenot, P. Pulkkinen, V. Kuhn, T. M. Link, F. Eckstein, andT. Jämsä, “Cortical bone finite element models in the estimation of experimentallymeasured failure loads in the proximal femur,” Bone 51, 737–40 (2012).

[82] J. Huopio, H. Kröger, R. Honkanen, J. S. Jurvelin, S. Saarikoski, and E. Alhava, “Cal-caneal ultrasound predicts early postmenopausal fractures as well as axial BMD. Aprospective study of 422 women,” Osteoporosis international : a journal established asresult of cooperation between the European Foundation for Osteoporosis and the National Os-teoporosis Foundation of the USA 15, 190–5 (2004).

[83] A. Stewart, V. Kumar, and D. M. Reid, “Long-term fracture prediction by DXA andQUS: a 10-year prospective study,” Journal of bone and mineral research : the officialjournal of the American Society for Bone and Mineral Research 21, 413–8 (2006).

[84] A. Moayyeri, S. Kaptoge, N. Dalzell, S. Bingham, R. N. Luben, N. J. Wareham, J. Reeve,and K. T. Khaw, “Is QUS or DXA better for predicting the 10-year absolute risk offracture?,” Journal of bone and mineral research : the official journal of the American Societyfor Bone and Mineral Research 24, 1319–25 (2009).

[85] A. Moayyeri, J. E. Adams, R. A. Adler, M.-A. Krieg, D. Hans, J. Compston, and E. M.Lewiecki, “Quantitative ultrasound of the heel and fracture risk assessment: an up-dated meta-analysis,” Osteoporosis international : a journal established as result of co-operation between the European Foundation for Osteoporosis and the National OsteoporosisFoundation of the USA 23, 143–53 (2012).


ii

ii

ii

ii

Bibliography

[86] K. A. Wear and B. S. Garra, “Assessment of bone density using ultrasonic backscatter,”Ultrasound in medicine & biology 24, 689–95 (1998).

[87] C. C. Glüer, C. Y. Wu, M. Jergas, S. A. Goldstein, and H. K. Genant, “Three quantitativeultrasound parameters reflect bone structure,” Calcified tissue international 55, 46–52(1994).

[88] O. Riekkinen, M. A. Hakulinen, M. J. Lammi, J. S. Jurvelin, A. S. Kallioniemi, andJ. Töyräs, “Acoustic properties of trabecular bone–relationships to tissue composi-tion,” Ultrasound in medicine & biology 33, 1438–44 (2007).

[89] O. Riekkinen, M. A. Hakulinen, J. Töyräs, and J. S. Jurvelin, “Spatial variation ofacoustic properties is related with mechanical properties of trabecular bone,” Physicsin medicine and biology 52, 6961–8 (2007).

[90] Q. Grimal, J. Grondin, S. Guérard, R. Barkmann, K. Engelke, C. C. Glüer, andP. Laugier, “Quantitative ultrasound of cortical bone in the femoral neck predictsfemur strength: results of a pilot study,” Journal of bone and mineral research : the officialjournal of the American Society for Bone and Mineral Research 28, 302–12 (2013).

[91] J. P. Karjalainen, J. Töyräs, O. Riekkinen, M. A. Hakulinen, and J. S. Jurvelin, “Ul-trasound backscatter imaging provides frequency-dependent information on struc-ture, composition and mechanical properties of human trabecular bone,” Ultrasoundin medicine & biology 35, 1376–84 (2009).

[92] J. D. Craven, M. A. Costantini, M. A. Greenfield, and R. Stern, “Measurement of thevelocity of ultrasound in human cortical bone and its potential clinical importance.An in vivo preliminary study,” Investigative radiology 8, 72–7 (1973).

[93] C. C. Glüer, R. Eastell, D. M. Reid, D. Felsenberg, C. Roux, R. Barkmann, W. Timm,T. Blenk, G. Armbrecht, A. Stewart, J. Clowes, F. E. Thomasius, and S. Kolta, “Associ-ation of five quantitative ultrasound devices and bone densitometry with osteoporoticvertebral fractures in a population-based sample: the OPUS Study,” Journal of bone andmineral research : the official journal of the American Society for Bone and Mineral Research19, 782–93 (2004).

[94] J. D. N. Cheeke, Fundamentals and Applications of Ultrasonic Waves (CRC Press LLC,Boca Raton, Florida, 2002).

[95] P. N. T. Wells, Physical principles of ultrasonic diagnosis (Academic Press Inc., London,New York, 1969).

[96] P. N. T. Wells, Biomedical Ultrasonics (Academic Press inc., London, 1977).

[97] C. F. Njeh, D. Hans, T. Fuerst, C. C. Glüer, and H. K. Genant, Quantitative Ultrasound:assessment of osteoporosis and bone status, 1st ed. (Martin Dunitz Ltd., London, 1999).

[98] L. Cardoso, F. Teboul, L. Sedel, C. Oddou, and A. Meunier, “In vitro acoustic wavespropagation in human and bovine cancellous bone,” Journal of bone and mineral research: the official journal of the American Society for Bone and Mineral Research 18, 1803–12(2003).

[99] Z. E. A. Fellah, J. Y. Chapelon, S. Berger, W. Lauriks, and C. Depollier, “Ultrasonicwave propagation in human cancellous bone: application of Biot theory,” The Journalof the Acoustical Society of America 116, 61–73 (2004).


ii

ii

ii

ii


[100] A. Hosokawa and T. Otani, “Ultrasonic wave propagation in bovine cancellous bone,”The Journal of the Acoustical Society of America 101, 558–62 (1997).

[101] A. Hosokawa and T. Otani, “Acoustic anisotropy in bovine cancellous bone,” TheJournal of the Acoustical Society of America 103, 2718–22 (1998).

[102] M. Pakula, F. Padilla, and P. Laugier, “Influence of the filling fluid on frequency-dependent velocity and attenuation in cancellous bones between 0.35 and 2.5 MHz,”The Journal of the Acoustical Society of America 126, 3301–10 (2009).

[103] K. Yamashita, F. Fujita, K. Mizuno, I. Mano, and M. Matsukawa, “Two-wave prop-agation imaging to evaluate the structure of cancellous bone,” IEEE transactions onultrasonics, ferroelectrics, and frequency control 59, 1160–6 (2012).

[104] K. Mizuno, H. Somiya, T. Kubo, M. Matsukawa, T. Otani, and T. Tsujimoto, “Influenceof cancellous bone microstructure on two ultrasonic wave propagations in bovinefemur: an in vitro study,” The Journal of the Acoustical Society of America 128, 3181–9(2010).

[105] K. A. Wear, “Measurements of phase valocity and group velocity in human calcaneus,”Ultrasound in medicine & biology 26, 641–646 (2000).

[106] K. A. Wear, “The dependencies of phase velocity and dispersion on trabecular thick-ness and spacing in trabecular bone-mimicking phantoms,” The Journal of the AcousticalSociety of America 118, 1186 (2005).

[107] K. A. Wear, “Group velocity, phase velocity, and dispersion in human calcaneus invivo,” The Journal of the Acoustical Society of America 121, 2431 (2007).

[108] K. A. Wear, “The dependencies of phase velocity and dispersion on volume fraction incancellous-bone-mimicking phantoms,” The Journal of the Acoustical Society of America125, 1197–201 (2009).

[109] G. Haïat, M. Sasso, S. Naili, and M. Matsukawa, “Ultrasonic velocity dispersion inbovine cortical bone: an experimental study,” The Journal of the Acoustical Society ofAmerica 124, 1811–21 (2008).

[110] B. Auld, Acoustic fields and waves in solids volume I, Second edi ed. (Krieger publishingcompany, Inc., Malabar, 1990).

[111] “CyberLogic Inc., Wave 2000 Pro manual,” (1999).

[112] K. Raum, R. O. Cleveland, F. Peyrin, and P. Laugier, “Derivation of elastic stiffnessfrom site-matched mineral density and acoustic impedance maps,” Physics in medicineand biology 51, 747–58 (2006).

[113] J. J. Faran, “Sound Scattering by Solid Cylinders and Spheres,” The Journal of theAcoustical Society of America 23, 405–418 (1951).

[114] K. A. Wear, F. Padilla, and P. Laugier, “Comparison of the Faran Cylinder Model andthe Weak Scattering Model for predicting the frequency dependence of backscatterfrom human cancellous femur in vitro,” The Journal of the Acoustical Society of America124, 1408–10 (2008).


ii

ii

ii

ii

Bibliography

[115] V. Kilappa, P. Moilanen, L. Xu, P. H. F. Nicholson, J. Timonen, and S. Cheng, “Low-frequency axial ultrasound velocity correlates with bone mineral density and corticalthickness in the radius and tibia in pre- and postmenopausal women,” Osteoporosisinternational : a journal established as result of cooperation between the European Foundationfor Osteoporosis and the National Osteoporosis Foundation of the USA 22, 1103–13 (2011).

[116] P. Moilanen, M. Määttä, V. Kilappa, L. Xu, P. H. F. Nicholson, M. Alén, J. Timonen,T. Jämsä, and S. Cheng, “Discrimination of fractures by low-frequency axial trans-mission ultrasound in postmenopausal females,” Osteoporosis international : a journalestablished as result of cooperation between the European Foundation for Osteoporosis and theNational Osteoporosis Foundation of the USA 24, 723–30 (2013).

[117] S. Bréban, F. Padilla, Y. Fujisawa, I. Mano, M. Matsukawa, C. L. Benhamou, T. Otani,P. Laugier, and C. Chappard, “Trabecular and cortical bone separately assessed at ra-dius with a new ultrasound device, in a young adult population with various physicalactivities,” Bone 46, 1620–5 (2010).

[118] C. M. Langton and C. F. Njeh, “The measurement of broadband ultrasonic attenua-tion in cancellous bone–a review of the science and technology,” IEEE transactions onultrasonics, ferroelectrics, and frequency control 55, 1546–54 (2008).

[119] C. F. Njeh, D. Hans, J. Li, B. Fan, T. Fuerst, Y. Q. He, E. Tsuda-Futami, Y. Lu, C. Y. Wu,and H. K. Genant, “Comparison of six calcaneal quantitative ultrasound devices: pre-cision and hip fracture discrimination,” Osteoporosis international : a journal establishedas result of cooperation between the European Foundation for Osteoporosis and the NationalOsteoporosis Foundation of the USA 11, 1051–62 (2000).

[120] K. A. Wear, “A method for improved standardization of in vivo calcaneal time-domainspeed-of-sound measurements,” IEEE transactions on ultrasonics, ferroelectrics, and fre-quency control 55, 1473–9 (2008).

[121] C. C. Glüer, “Quantitative Ultrasound–it is time to focus research efforts,” Bone 40,9–13 (2007).

[122] B. K. Hoffmeister, A. R. Wilson, M. J. Gilbert, and M. E. Sellers, “A backscatter differ-ence technique for ultrasonic bone assessment,” The Journal of the Acoustical Society ofAmerica 132, 4069–76 (2012).

[123] E. Bossy, M. Talmant, F. Peyrin, L. Akrout, P. Cloetens, and P. Laugier, “An in vitrostudy of the ultrasonic axial transmission technique at the radius: 1-MHz velocitymeasurements are sensitive to both mineralization and intracortical porosity,” Journalof bone and mineral research : the official journal of the American Society for Bone and MineralResearch 19, 1548–56 (2004).

[124] P. Moilanen, “Ultrasonic guided waves in bone,” IEEE Trans Ultrason Ferroelectr FreqControl 55, 1277–1286 (2008).

[125] M. Muller, D. Mitton, P. Moilanen, V. Bousson, M. Talmant, and P. Laugier, “Predic-tion of bone mechanical properties using QUS and pQCT: study of the human distalradius,” Medical engineering & physics 30, 761–7 (2008).

[126] R. Barkmann, P. Laugier, U. Moser, S. Dencks, F. Padilla, G. Haïat, M. Heller, and C. C.Glüer, “A method for the estimation of femoral bone mineral density from variablesof ultrasound transmission through the human femur,” Bone 40, 37–44 (2007).


ii

ii

ii

ii


[127] R. Barkmann, P. Laugier, U. Moser, S. Dencks, M. Klausner, F. Padilla, G. Haïat, andC. C. Glüer, “A device for in vivo measurements of quantitative ultrasound variablesat the human proximal femur,” IEEE Trans Ultrason Ferroelectr Freq Control 55, 1197–1204 (2008).

[128] R. Barkmann, S. Dencks, P. Laugier, F. Padilla, K. Brixen, J. Ryg, A. Seekamp,L. Mahlke, A. Bremer, M. Heller, and C. C. Glüer, “Femur ultrasound (FemUS)–first clinical results on hip fracture discrimination and estimation of femoral BMD,”Osteoporosis international : a journal established as result of cooperation between the Euro-pean Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 21,969–76 (2010).

[129] S. Dencks, R. Barkmann, F. Padilla, G. Haïat, P. Laugier, and C. C. Glüer, “Wavelet-based signal processing of in vitro ultrasonic measurements at the proximal femur,”Ultrasound in medicine & biology 33, 970–80 (2007).

[130] S. Dencks, R. Barkmann, F. Padilla, P. Laugier, G. Schmitz, and C. C. Glüer, “Model-based estimation of quantitative ultrasound variables at the proximal femur,” IEEEtransactions on ultrasonics, ferroelectrics, and frequency control 55, 1304–15 (2008).

[131] G. Haïat, F. Padilla, R. Barkmann, S. Dencks, U. Moser, C. C. Glüer, and P. Laugier,“Optimal prediction of bone mineral density with ultrasonic measurements in excisedhuman femur,” Calcified tissue international 77, 186–92 (2005).

[132] G. Haïat, F. Padilla, R. O. Cleveland, and P. Laugier, “Effects of frequency-dependentattenuation and velocity dispersion on in vitro ultrasound velocity measurements inintact human femur specimens,” IEEE transactions on ultrasonics, ferroelectrics, and fre-quency control 53, 39–51 (2006).

[133] F. Padilla, L. Akrout, S. Kolta, C. Latremouille, C. Roux, and P. Laugier, “In vitroultrasound measurement at the human femur,” Calcified tissue international 75, 421–30(2004).

[134] P. Moilanen, Z. Zhao, P. Karppinen, T. Karppinen, V. Kilappa, J. Pirhonen, R. Myllylä,E. Haeggström, and J. Timonen, “Photo-acoustic excitation and optical detection offundamental flexural guided wave in coated bone phantoms,” Ultrasound in medicine& biology 40, 521–31 (2014).

[135] P. Moilanen, M. Talmant, V. Kilappa, P. Nicholson, S. Cheng, J. Timonen, andP. Laugier, “Modeling the impact of soft tissue on axial transmission measurementsof ultrasonic guided waves in human radius,” J Acoust Soc Am 124, 2364–2373 (2008).

[136] J. P. Karjalainen, O. Riekkinen, J. Töyräs, H. Kröger, J. S. Jurvelin, J. Töyräs, andH. Kröger, “Ultrasonic assessment of cortical bone thickness in vitro and in vivo,”IEEE Trans Ultrason Ferroelectr Freq Control 55, 2191–2197 (2008).

[137] J. P. Karjalainen, O. Riekkinen, J. Töyräs, M. A. Hakulinen, H. Kröger, T. Rikkonen,K. Salovaara, and J. S. Jurvelin, “Multi-site bone ultrasound measurements in elderlywomen with and without previous hip fractures,” Osteoporosis international : a journalestablished as result of cooperation between the European Foundation for Osteoporosis and theNational Osteoporosis Foundation of the USA 23, 1287–95 (2012).

[138] B. K. Hoffmeister, S. A. Whitten, S. C. Kaste, and J. Y. Rho, “Effect of collagen and min-eral content on the high-frequency ultrasonic properties of human cancellous bone,”Osteoporosis international : a journal established as result of cooperation between the Euro-pean Foundation for Osteoporosis and the National Osteoporosis Foundation of the USA 13,26–32 (2002).


ii

ii

ii

ii

Bibliography

[139] M. A. Hakulinen, J. Töyräs, S. Saarakkala, J. Hirvonen, H. Kröger, and J. S. Jurvelin,“Ability of ultrasound backscattering to predict mechanical properties of bovine tra-becular bone,” Ultrasound in medicine & biology 30, 919–27 (2004).

[140] M. A. Hakulinen, J. S. Day, J. Töyräs, M. Timonen, H. Kröger, H. Weinans, I. Kiviranta,and J. S. Jurvelin, “Prediction of density and mechanical properties of human trabecu-lar bone in vitro by using ultrasound transmission and backscattering measurementsat 0.2-6.7 MHz frequency range,” Phys Med Biol 50, 1629–1642 (2005).

[141] M. A. Hakulinen, J. S. Day, J. Töyräs, H. Weinans, and J. S. Jurvelin, “Ultrasoniccharacterization of human trabecular bone microstructure,” Physics in medicine andbiology 51, 1633–48 (2006).

[142] B. K. Hoffmeister, C. I. Jones 3rd, G. J. Caldwell, and S. C. Kaste, “Ultrasonic char-acterization of cancellous bone using apparent integrated backscatter,” Phys Med Biol51, 2715–2727 (2006).

[143] B. K. Hoffmeister, D. P. Johnson, J. A. Janeski, D. A. Keedy, B. W. Steinert, A. M. Viano,and S. C. Kaste, “Ultrasonic characterization of human cancellous bone in vitro usingthree different apparent backscatter parameters in the frequency range 0.6-15.0 MHz,”IEEE Trans Ultrason Ferroelectr Freq Control 55, 1442–1452 (2008).

[144] B. K. Hoffmeister, “Frequency dependence of apparent ultrasonic backscatter fromhuman cancellous bone,” Physics in medicine and biology 56, 667–83 (2011).

[145] C. Liu, D. Ta, F. Fujita, T. Hachiken, M. Matsukawa, K. Mizuno, and W. Wang, “Therelationship between ultrasonic backscatter and trabecular anisotropic microstructurein cancellous bone,” Journal of Applied Physics 115, 064906 (2014).

[146] R. Kuc, “Clinical application of an ultrasound attenuation coefficient estimation tech-nique for liver pathology characterization,” IEEE transactions on bio-medical engineering27, 312–9 (1980).

[147] J. G. Miller, J. E. Pérez, and B. E. Sobel, “Ultrasonic characterization of myocardium,”Progress in cardiovascular diseases 28, 85–110 (1985).

[148] F. L. Lizzi, M. Ostromogilsky, E. J. Feleppa, M. C. Rorke, and M. M. Yaremko, “Rela-tionship of ultrasonic spectral parameters to features of tissue microstructure,” IEEEtransactions on ultrasonics, ferroelectrics, and frequency control 34, 319–29 (1987).

[149] J. Y. Rho, P. Zioupos, J. D. Currey, and G. M. Pharr, “Microstructural elasticity andregional heterogeneity in human femoral bone of various ages examined by nano-indentation,” J Biomech 35, 189–198 (2002).

[150] T. Hofmann, F. Heyroth, H. Meinhard, W. Franzel, and K. Raum, “Assessment of com-position and anisotropic elastic properties of secondary osteon lamellae,” J Biomech 39,2282–2294 (2006).

[151] M. L. Oyen, “Nanoindentation hardness of mineralized tissues,” Journal of biomechanics39, 2699–702 (2006).

[152] M. Kawabe, K. Fukui, M. Matsukawa, M. Granke, A. Saïed, Q. Grimal, and P. Laugier,“Comparative investigation of elastic properties in a trabecula using micro-Brillouinscattering and scanning acoustic microscopy,” The Journal of the Acoustical Society ofAmerica 132, EL54–60 (2012).


ii

ii

ii

ii


[153] M. Matsukawa, R. Tsubota, M. Kawabe, and K. Fukui, “Application of a micro-Brillouin scattering technique to characterize bone in the GHz range,” Ultrasonics 54,1155–61 (2014).

[154] Q. Grimal, K. Raum, A. Gerisch, and P. Laugier, “Derivation of the mesoscopic elas-ticity tensor of cortical bone from quantitative impedance images at the micron scale,”Comput Methods Biomech Biomed Engin 11, 147–157 (2008).

[155] K. Raum, “Microelastic imaging of bone,” IEEE Trans Ultrason Ferroelectr Freq Control55, 1417–1431 (2008).

[156] K. Raum, J. Reisshauer, and J. Brandt, “Frequency and resolution dependence ofthe anisotropic impedance estimation in cortical bone using time-resolved scanningacoustic microscopy,” J Biomed Mater Res A 71, 430–438 (2004).

[157] M. Granke, A. Gourrier, F. Rupin, K. Raum, F. Peyrin, M. Burghammer, A. Saïed, andP. Laugier, “Microfibril orientation dominates the microelastic properties of humanbone tissue at the lamellar length scale,” PloS one 8, e58043 (2013).

[158] J. S. Jurvelin, O. Riekkinen, and M. A. Hakulinen, “Method for measuring of thick-nesses of materials using an ultrasound technique,” (2010).

[159] E. Bossy, F. Padilla, F. Peyrin, and P. Laugier, “Three-dimensional simulation of ul-trasound propagation through trabecular bone structures measured by synchrotronmicrotomography,” Physics in medicine and biology 50, 5545–56 (2005).

[160] A. Hosokawa, “Simulation of ultrasound propagation through bovine cancellousbone using elastic and Biot’s finite-difference time-domain methods,” The Journal ofthe Acoustical Society of America 118, 1782 (2005).

[161] Y. Nagatani, H. Imaizumi, T. Fukuda, M. Matsukawa, Y. Watanabe, and T. Otani,“Applicability of Finite-Difference Time-Domain Method to Simulation of Wave Prop-agation in Cancellous Bone,” Japanese Journal of Applied Physics 45, 7186–7190 (2006).

[162] G. Haïat, F. Padilla, F. Peyrin, and P. Laugier, “Variation of ultrasonic parameters withmicrostructure and material properties of trabecular bone: a 3D model simulation,”Journal of bone and mineral research : the official journal of the American Society for Bone andMineral Research 22, 665–74 (2007).

[163] G. Haïat, F. Padilla, M. Svrcekova, Y. Chevalier, D. Pahr, F. Peyrin, P. Laugier, andP. Zysset, “Relationship between ultrasonic parameters and apparent trabecular boneelastic modulus: a numerical approach,” Journal of biomechanics 42, 2033–9 (2009).

[164] R. S. Schechter, H. H. Chaskelis, R. B. Mignogna, and P. P. Delsanto, “Real-timeparallel computation and visualization of ultrasonic pulses in solids,” Science (NewYork, N.Y.) 265, 1188–92 (1994).

[165] K. F. Graff, Wave motion in elastic solids (Oxford University Press, Oxford, 1975).

[166] J. J. Kaufman, G. Luo, and R. S. Siffert, “Ultrasound simulation for 3D-axisymmetricmodels,” in Ultrasonics, 2003 IEEE Symposium on Ultrasonics, 2003 IEEE Symposium onVO - 2, Vol. 2 (2003), pp. 2065–2068 Vol.2.

[167] O. Riekkinen, M. A. Hakulinen, M. Timonen, J. Töyräs, J. S. Jurvelin, and J. Töyräs,“Influence of overlying soft tissues on trabecular bone acoustic measurement at vari-ous ultrasound frequencies,” Ultrasound in medicine & biology 32, 1073–83 (2006).


ii

ii

ii

ii

Bibliography

[168] K. Raum, I. Leguerney, F. Chandelier, M. Talmant, A. Saïed, F. Peyrin, and P. Laugier,“Site-matched assessment of structural and tissue properties of cortical bone usingscanning acoustic microscopy and synchrotron radiation muCT,” Phys Med Biol 51,733–746 (2006).

[169] S. Lakshmanan, A. Bodi, and K. Raum, “Assessment of anisotropic tissue elasticityof cortical bone from high-resolution, angular acoustic measurements,” IEEE TransUltrason Ferroelectr Freq Control 54, 1560–1570 (2007).

[170] R. Hube, H. Mayr, W. Hein, and K. Raum, “Prediction of biomechanical stabilityafter callus distraction by high resolution scanning acoustic microscopy,” Ultrasoundin medicine & biology 32, 1913–21 (2006).

[171] B. Preininger, S. Checa, F. L. Molnar, P. Fratzl, G. N. Duda, and K. Raum, “Spatial-temporal mapping of bone structural and elastic properties in a sheep model followingosteotomy,” Ultrasound in medicine & biology 37, 474–83 (2011).

[172] C. C. Glüer, G. Blake, Y. Lu, B. A. Blunt, M. Jergas, and H. K. Genant, “Accurateassessment of precision errors: how to measure the reproducibility of bone densitom-etry techniques,” Osteoporosis international : a journal established as result of cooperationbetween the European Foundation for Osteoporosis and the National Osteoporosis Foundationof the USA 5, 262–70 (1995).

[173] M. Gomez, F. Aguado, J. Manuel, J. M. Menendez, M. Revilla, L. F. Villa, J. Cortes,H. Rico, M. Gómez, J. M. Menéndez, and J. Cortés, “Influence of soft tissue (fat andfat-free mass) on ultrasound bone velocity: an in vivo study,” Investigative radiology32, 609–12 (1997).

[174] P. O. Kotzki, D. Buyck, D. Hans, E. Thomas, F. Bonnel, F. Favier, P. J. Meunier, andM. Rossi, “Influence of fat on ultrasound measurements of the os calcis,” Calcifiedtissue international 54, 91–5 (1994).

[175] C. Chappard, E. Camus, F. Lefebvre, G. Guillot, J. Bittoun, G. Berger, and P. Laugier,“Evaluation of error bounds on calcaneal speed of sound caused by surrounding softtissue,” Journal of clinical densitometry : the official journal of the International Society forClinical Densitometry 3, 121–31 (2000).

[176] M. S. Stein, S. A. Feik, C. D. L. Thomas, J. G. Clement, and J. D. Wark, “An automatedanalysis of intracortical porosity in human femoral bone across age,” Journal of boneand mineral research : the official journal of the American Society for Bone and MineralResearch 14, 624–32 (1999).

[177] V. Bousson, C. Bergot, A. Meunier, F. Barbot, C. Parlier-Cuau, A. M. Laval-Jeantet,and J. D. Laredo, “CT of the middiaphyseal femur: cortical bone mineral density andrelation to porosity,” Radiology 217, 179–87 (2000).

[178] X. Wang and Q. Ni, “Determination of cortical bone porosity and pore size distribu-tion using a low field pulsed NMR approach,” Journal of orthopaedic research : officialpublication of the Orthopaedic Research Society 21, 312–9 (2003).

[179] C. D. L. Thomas, S. A. Feik, and J. G. Clement, “Regional variation of intracorticalporosity in the midshaft of the human femur: age and sex differences,” Journal ofanatomy 206, 115–25 (2005).

[180] J. C. Wall, S. K. Chatterji, and J. W. Jeffery, “Age-related changes in the density andtensile strength of human femoral cortical bone,” Calcif Tissue Int 27, 105–108 (1979).


ii

ii

ii

ii


[181] D. Rohrbach, S. Lakshmanan, F. Peyrin, M. Langer, A. Gerisch, Q. Grimal, P. Laugier,and K. Raum, “Spatial distribution of tissue level properties in a human femoralcortical bone,” J Biomech 45, 2264–2270 (2012).

[182] G. Bergmann, G. Deuretzbacher, M. Heller, F. Graichen, A. Rohlmann, J. Strauss, andG. N. Duda, “Hip contact forces and gait patterns from routine activities,” J Biomech34, 859–871 (2001).

[183] C. H. Turner, J. Rho, Y. Takano, T. Y. Tsui, and G. M. Pharr, “The elastic proper-ties of trabecular and cortical bone tissues are similar: results from two microscopicmeasurement techniques,” J Biomech 32, 437–441 (1999).

[184] P. Guggenbuhl, “Osteoporosis in males and females: Is there really a difference?,”Joint, bone, spine : revue du rhumatisme 76, 595–601 (2009).

[185] M. J. Turunen, V. Prantner, J. S. Jurvelin, H. Kröger, and H. Isaksson, “Compositionand microarchitecture of human trabecular bone change with age and differ betweenanatomical locations,” Bone 54, 118–25 (2013).

[186] S. Chaffaî, F. Peyrin, S. Nuzzo, R. Porcher, G. Berger, and P. Laugier, “Ultrasoniccharacterization of human cancellous bone using transmission and backscatter mea-surements: relationships to density and microstructure,” Bone 30, 229–37 (2002).

[187] F. Jenson, F. Padilla, V. Bousson, C. Bergot, J. D. Laredo, and P. Laugier, “In vitroultrasonic characterization of human cancellous femoral bone using transmission andbackscatter measurements: Relationships to bone mineral density,” The Journal of theAcoustical Society of America 119, 654 (2006).

[188] K. A. Wear and D. W. Armstrong, “Relationships among calcaneal backscatter, atten-uation, sound speed, hip bone mineral density, and age in normal adult women,” TheJournal of the Acoustical Society of America 110, 573 (2001).

[189] F. Padilla, F. Jenson, V. Bousson, F. Peyrin, and P. Laugier, “Relationships of trabecularbone structure with quantitative ultrasound parameters: in vitro study on humanproximal femur using transmission and backscatter measurements,” Bone 42, 1193–202 (2008).

[190] K. A. Wear, S. Nagaraja, M. L. Dreher, and S. L. Gibson, “Relationships of quantitativeultrasound parameters with cancellous bone microstructure in human calcaneus invitro,” The Journal of the Acoustical Society of America 131, 1605–12 (2012).

[191] B. K. Hoffmeister, A. Holt, and S. Kaste, “Effect of the cortex on ultrasonic backscattermeasurements of cancellous bone,” Physics in medicine and . . . 56, 6243–6255 (2011).

[192] J. L. Rose, Ultrasonic Waves in Solid Media (Cambridge University Press, 2004).

[193] C. C. Glüer, “A new quality of bone ultrasound research,” IEEE Trans Ultrason Ferro-electr Freq Control 55, 1524–1528 (2008).

[194] D. Hans and M. A. Krieg, “Quantitative ultrasound for the detection and managementof osteoporosis,” Salud Publica Mex 51 Suppl 1, S25–37 (2009).


Publications of the University of Eastern Finland Dissertations in Forestry and Natural Sciences

Publications of the University of Eastern Finland

Dissertations in Forestry and Natural Sciences

isbn 978-952-61-1530-6

Markus Malo

Quantitative Characterization of Proximal Femur Using Pulse-Echo Ultrasound Measurements

Osteoporosis is a common bone

disease leading to increased fragility

and fracture probability. However,

only quarter of individuals suffering

from the disease have received a

diagnosis. For effective management

of the disease it would be highly

important to develop diagnostic

tools capable of mass screening of

the population at the basic level of

healthcare. In this thesis quantitative

pulse-echo ultrasound technique for

evaluation of proximal femur was

investigated and developed towards

this goal by means of numerical

modelling and in vitro and ex vivo

measurements.

dissertatio

ns | 145 | M

ar

ku

s Ma

lo | Q

ua

ntita

tive Ch

ara

cterizatio

n of P

roxim

al F

emu

r Usin

g Pu

lse-Ech

o U

ltraso

un

d...

Markus MaloQuantitative

Characterization of Proximal Femur Using Pulse-Echo Ultrasound

Measurements

| 145 | Markus Malo | Quantitative Characterization of Proximal … · Proximal Femur Using Pulse-Echo Ultrasound Measurements Osteoporosis is a common bone disease leading to increased

Documents