-
DEVELOPMENT OF A HAND TREMOR QUANTIFICATION DEVICE FOR THE
MEASUREMENT OF PATHOLOGICAL TREMOR
Sonja Markez
A thesis submitted in conformity with the requirements for the
degree of Master of Health Science,
Institute of Biomaterids and Biomedical Engineering, University
of Toronto
O Copyright by Sonja Markez 2000
-
National Library mJ)Im of Canada Bibliothèque nationale du
Canada
Acquisitions and Acquisitions et Bibliographie Services services
bibliographiques 395 Wellington Street 395, rue Wellington OttawaON
KIAON4 Ottawa ON K I A ON4 Canada Canada
The author bas granted a non- exclusive licence allowing the
National Library of Canada to reproduce, loan, distribute or sell
copies of this thesis in microforrn, paper or electronic
formats.
L'auteur a accordé une licence non exclusive permettant à la
Bibliothèque nationale du Canada de reproduire, prêter, distribuer
ou vendre des copies de cette thèse sous la forme de microfiche/fh,
de reproduction sur papier ou sur format électronique.
The author retains ownership of the L'auteur conserve la
propriété du copyright in this thesis. Neither the droit d'auteur
qui protège cette thèse. thesis nor substantial extracts tiom it Ni
la thèse ni des extraits substantiels may be printed or othewise de
celle-ci ne doivent être imprimés reproduced without the author's
ou autrement reproduits sans son permission. autorisation.
-
Development of a Hand Tremor Quantification Device for the
Measurernent of Pathologieal Tremor
Sonja Markez
Institute of Biomateriais and Biomedical Engineering University
of Toronto
2000
ABSTRACT
Hand tremor is the most cornmon and visible symptom of a variety
of neurological
disordea, including Parkinson's disease. Consequently, the
seventy of tremor is oflen
used as an aid in diagnosis and a gauge to assess the eficacy of
treatment.
A hand tremor quantification device, the miniBIRDTM/Tremor
Quantifer system, was
designed with the electromagnetic tracker, miniBIRDm fiom
Ascension Technology, and
customized software. The miniBtRDm's sensor is affixed to the
finger and data are
acquired through a host computer. The system is set to rneasure
the three main types of
tremor associated with Parkinson's disease: resting, postural,
and intention. Frequency,
amplitude, and area of displacement of tremor (an indication of
power) are determined
for each set of data,
The system was tested to a positional accuracy of (0.031 +
0.132) cm averaged over a specified range. The fiequency
measurements were tested against those an
accelerometer and found to be comparable.
Development of a Hand Tremor Quantincation Device for the
Measraement of Pathological Tremor ii
-
The timely completion of this project memt a great deal to me
and I appreciate the
contribution of those who helped to make it possible. The
following acknowledges their
contributions but 1 hope that my personal expressions of thanks
have already done so.
Thank you to Dr. Tony Easty and Joe C d ~ o who gave me their
tirne, support, and
guidance. Tony, 1 value the trust that you implied with your
patience and support. Joe, it
has been my privilege to work under your supervision for al1
these years.
Thank you to Jack Lam whose assistance cannot be quantified. 1
think that 1 have leamed
more 'engineering' from you than I have in six years of
univenity.
1 am fortunate to have been surrounded by good enginees and
colleagues.
Many thanks to:
Rich Leask for his help, advice, and technical support. Th
appreciate his kindness nonetheless.
ough it is his nature. 1
Rob Ebetsch for his proof-reading, sense of humour, and being a
great person to talk to.
Des Campbell for his fme craftsmanship and fnendly smile.
John Leung for his generosity in Iending me his digital
carnera.
Dave Brewin for not steaiing my chair.. . and for al1 his
logistical help.
Katina Di Biase for her humour and niendship.
Thank you to my colleagues Sonia Pinkney and Vicky Young,
without whose fnendship I
surely would have jumped ship a long t h e ago.
Most of all, thank you to my family: John, Frank, and Alenka
Markez. Needless to Say
there would be no Master's if it were not for your patience and
support.
Devetopment of a Hand Tremor Quantification Device for the
Meamment of Pathologicai Tremor iii
-
TABLE OF CONTENTS
..................................................................................
ABSTRACT
................................................................
ACKNû WLEDGEMENTS
...................................................................
TABLE OF CONTENTS
........................................................................
LIST OF FIGURES
LIST OF TABLES
...........................................................................
............ ......................... Chapter 1: INTRODUCTION
......................
............................................................
Chapter 2: BACKGROUND
............................................................. . 2 t
Parkinson's Disease
............................................................ 2.1.1
Symptoms
2.1.1.1 Resting Tremor
.......................................... 2.1.1.2 Postural Tremor
.......................................... 2.1 . 1.3 Intention
Tremor ......................................... 2.1.1.4 Cogwheel
Phenomenon .................................
............................................................
2.1.2 Treatment 2.1.2.1 Anticholinergic Drugs
................................... 2.1 2 .2 Antidepressants
.......................................... 2.1.2.3 Levodopa
(L-dopa) ...................................... 2.1.3.4 Sinemet
.................................................... 2.1 .2.5
Dopaminergic Agonists ................................. 3.1 .3.6
Beta-Adrenergic Blocken .............................. 2.1.2.7
Thalamotomy .............................................
...................................................... 2.2
Classification of Tremor
...............................................................
Etiology
Amplitude. Waveforrn. Frequency ................................
Pharmac olo gicd Response ........................................
Appearance and Behavioural C haract e ristics
....................
2.2.4.1 Resting Trernor
.......................................... 2 . 2.4.2 Action Tremor
............................................ 2.2.4.3 Postural
Tremor .......................................... 2.2.4.4 Kinetic
Tremor ........................................... 2.2.4.5
Intention Tremor ......................................... 2.2.4.6
Task-Specinc or Occupational Kinetic Tremor ...... 2.2.4.7
Isometric Tremor ......................................... 2.2.4.8
Essential Tremor ......................................... 2.2.4.9
Static Tremor .............................................
................................. Chapter 3: MEASUREMENT OF
TREMOR ....
........................................................... 3.1
Review o f Literature
............................................. 3.1.1 Subjective
Assessrnent 3.1.1.1 Tremor Rating Scdes
....................................
.............................................. 3.1.2 Objective
Assessrnent 3.1.2.1 Tambour
...................................................
.......................................... 3.1.2.2 Optical
Methods 3.1.2.3 Accelerornetry
........................................... 3.1.2.4 Digitizing
Tabtets .......................................
Page . . 11
iii iv vi ... vlrl
1 3 3 4 4 6 7 7 8 8 9 9 10 10 10 10 11 12 12 13 14 14 14 14 15
15 15 15 15 15 17 17 17 17 20 20 20 21 22
Development of a IIand Tremor Quantification Device for the
Measurement of Pathologicai Tremor iv
-
3.1.2.5 Optoelectric Imaging Systems .........................
3.1.2.6 Surface Electromyography ..............................
3.2 Tremor C haracteristics
......................................................... 3.2.1
Wavefonn
............................................................
............................................................ 3 2
.2 Frequency 3.2.3 Amplitude
............................................................ 3.2.4
Long-Term Tremor Measurement ................................
3.3 Selection of a Measurement Device
......................................... 3.3.1 RingMouse
.......................................................... 3.3.2
Accelerometers ........................ ..
............................ 3.3.3 Electromagnetic Tracking
..........................................
Chapter 4: METHODOLOGY
.......................................................... 4.1
Description of Technology: m i n i B I R D T M
...................................
.................................................................
4.2 Serial Interface 4.3 Hardware Configuration
...................................................... 1.4 Software
Description
..........................................................
................................................... Chapter 5:
SYSTEM EVALUTION
................................................... 5.1 System
Evaluation Resuits 5.1.1 Test of Positional Accuracy
.......................................
5.1.1.1 Position Accuracy Test - Part One .....................
.................... 5.1.1.2 Position Accuracy Test - Part Two
5.1.2 Test of Frequency Measurement
.................................. 5.1.2.1 Frequency Measurement
Test - Part One ............. 5.1 2 .2 Frequency Measurement Test -
Part Two ............
................................. Chapter 6: DISCUSSION AND
CONCLUSIONS
.....................................................................
6.1 Introduction
...............................................................
6.2 System Strengths . .
............................................................ 6.3 S
ystem Limitations
............................................................... 6.4
Tremor Examples
............................................................. 6.4 C
M c d Relevance
6.5 Conclusions
.....................................................................
6.6 Future Work
.....................................................................
.......................................................................................
Glossary
....................................................................................
References .... Appendix A: Hoehn and Yahr Scale of Clinical Stages
of Parkinson's Disease
Appendix B: miniBIRDm Specifications .......... ....
.............................. Appendix C : FASTRAKfM S
pecifications Compared to rniniBIRDTM ............... Appendix D: S
tatic Position Test Data - Part One
...................................... Appenduc E: Static Position
Test Data - Part Two ......................................
Appendix F: Entran@ Accelerometer Specincations - Model EGAX-F
............ Appendix G: Frequency Test Data With 'Oscillator'.
...................................
Development of a Hand Tremor Quantification Device for the
Mea~ufement of Pathological T m o r v
-
LIST OF FIGURES
Figure 2.1 Figure 3.1 Figure 3.2 Figure 3.3 Figure 3.4 Figure
3.5 Fi-me 3.6 Figure 4.1 Figure 4.2 Figure 4.3 Figure 4.4 Figure
4.5 Figure 4.6 Figure 4.7 Figure 4.8 Figure 1.9
Figure 4.10
Figure 4.1 1
Figure 4.12 Figure 4.1 3
Figure 4.14
Figure 4.1 5
Figure 4.16 Figure 4.17 Figure 4.18 Figure 4.19 Figure 4.20
Figure 4.21 Figure 4.22 Figure 4.23
Figure 4.24 Figure 4.25 Figure 4.26 Figure 5.1 Figure 5.2 Figure
5.3
Tremor classification by appearance and behavioural
characteristics .
.......................................................... Pegasus
RingMouse
Close-up of Entran EGAX-F uniaxial accelerometer
.................... miniBIRDTM h m Ascension Technology
................................ FASTRAKTM fiom Poihemus Inc
.......................................... miniBIRDTM versus
FASTRAKfM .......................................... miniBIRDm
sensor: mode! 800 versus model 500 ...................... Screen
shot of Tremor @anrifer ...........................................
Movement of hand in x direction of intention test
........................ Movement of hand in y direction of
intention test ........................ Movement of hand in z
direction of intention test ....................... LabVIEWTM
sequence used to subtract mean fiom data ................ Movement
of hand in x direction of intention test centred about zero
Movement of hand in y direction of intention test centred about
zero Movement of hand in z direction of intention test centred
about zero Movement of hand in x direction of intention test
centred about zero
......................................................... and
hi&-pass filtered Movement of hand in y direction of intention
test centred about zero
......................................................... and
hi&-pass filtered Movement of hand in z direction of intention
test centred about zero
........................................................ and
high-pass filtered ........................... LabVIEWTM Auto Power
Spectnim fûnction
X direction motion of sample tremor and corresponding power
........................................................................
spectnim
Y direction motion of sample tremor and corresponding power
spectnun
........................................................................
Z direction motion of sarnple tremor and corresponding power
spectrum
........................................................................
.................... LabVIEWRn Power & Frequency Estimate
function ....................................... Deviation From
Mean @FM) c w e
............... Deviation From Mean (DFM) curve for intention
tremor ...................... High-pass Ntered DFM cuve for
intention tremor
.................. DFM curve with area of displacement of tremor
of 6.9
................. DFM curve with area of displacement of tremor
of 1.4 ................... Deviation From Mean @FM) plot for sample
tremor
Signed Deviation From Mean (SDFM) plot for the same tremor
displayed in Figure 4.22
......................................................
................... SDFM curve and accornpanying frequency
spectrum ..................... DFM curve and accompanying kquency
spectnim
.................................. Sample printout fiom Tremor
Quuntij?er .......................................... Plexiglass
grid: position test setup
..................... miniBIRDfM sensor in plexiglass position
test sehip ... Close-up of plexiglass position test setup with
miniBIRDm sensor
DeveIopment of a Hmd Tremor Quantification Device for the
Measurement of Paihological Tremor vi
-
Figure 5.4
Figure 5.5 Figure 5.6
Figure 5.7 Figure 5.8 Figure 5.9 Figure 5.10 Figure 6.1 Figure
6.2 Figure 6.3 Figure 6.4
Cornparison of tolerance of plexiglass grid and miniBIRDTM
pubfished error
................................................................
Distribution of position error
................................................ Mean error and
standard deviation of measurement readings fiom
...............................................................
the miniBIRDTM Entran accelerometer next to d I R D m sensor
........................ Frequency test setup with accelerometer
and miniBIRDa sensor ..... Accelerometer and rniniBIRDTM sensor on
tip of 'oscillator'. ...........
............. m i n i B I R D T M sensor and accelerometer
mounted on &ger
............. Tremor Quantifier screen shot of simulated resting
tremor ............ Tremor Banrifer screen shot of simulated
postural tremor .......... Tremor Quantifier screen shot of
simulated intention tremor
Tremor Quantijer screen shot of simulated intention tremor -
SDFM high-pass filtered
...................................................
Devdopment of a Hand Tremor Quantification Device for the
Meamremnt of Pathological Tremor vii
-
LIST OF TABLES
Table 2.1
Table 3.1 Table 3.2 Table 3.3 Table 5.1 Table 5.2 Table 5.3
Table 5.4
Table 5.5 Table 5.6
Table 5.7 Table 6.1
Classification of common tremors by fiequency and behavioural
.....................................................................
appearance
Webster Scde for parkinsonian tremor
.................................... ............................
Unifïed Rating Scale for parkinsonian fremor
........................................... Rathg scale for
essential tremor Sarnple static position test data - Part one
................................. Descriptive statistics on
position enor ..................................... Enor associated
with measurement of various distances ................ Frequency
measurements ushg miniBIRDTM and accelerometer on
.......................................................................
'oscillator' ....................... Pair t-test results for
frequency test on 'oscillator'.
Frequency measurements using miniBIRDm and accelerometer on
...........................................................................
finger
.......................... Paired t-test results for fiequency
test on h g e r Strengths and limitations of miniBIRDm/Tremor
Quantifier system
Page
13 18 18 19 70 72 76
82 83
85 85 89
Development of a Hand Tremor Quantification Device for the
Measurement of Paihological Tremor viii
-
Chapter 1 Introduction
Chapter 1 INTRODUCTION
Hand tremor is the most common and visible symptom of
Parkinson's disease. As a
result, its severity is ofien used as an indication of the
ihess' stage of progression. It cm
aiso be used to assess the efncacy of phamiacological or
surgical treatment.
Attempts to quanti@ the characteristics of hand tremor have
occupied researchen for
over a century and although there are a number of systems that
can quantiq tremor, each
saers fiom its own shortcomings whether it be prohibitive cost,
insunicient accuracy,
unwieldy hardware, etc.
In 1989, Dr. Peter Ashby a neurologist at the University Health
~etwork ' approached Dr.
Tony Easty, Director of Medical Engineering, requesting a
low-cost systern that could be
used to measure hand tremor in Parkinson's patients. The project
was taken on by
Elizabeth Sheil, a graduate student at the institute of
Biomatenals and Biomedical
~ng ineer in~~ , University of Toronto. Her Master's thesis,
entitied 771e Quantification of
Hmd Tremor in Clinical Newological Assessment (1991), descnbes
the construction of a
video imaging system that used CCD cameras to track the motion
of an ÙiFared LED.
This motion was then analyzed with custom-written software. The
system was M e r
improved by Joseph Cafazzo in his Master's thesis: Development
of a Hand Tremor
Quantiifier for Chical Neurological Assessment ( 1 992).
Although the video imagîng system fÙI.filled al1 technical
requirements it was comprised
of a large amount of bulky hardware including two CCD cameras,
two moniton, a large
1 Toronto Western Hospital ï hen caiied the Iastitute of
Biomedical Engineering
Development of a Hand Tremor Quantification Device for the
Mea~urement of Pathological Tremor 1
-
Chapter I Introduction
box area for testing, and a cornputer. As a result of its
relative non-portability it was
never implemented clinically.
This project is an attempt to revisit this clinical need by
developing a relatively low-cost,
portable hand tremor quantification device that can measure
pathological tremor.
Although the project specifications are identical to those of
the previoiis two atternpts, the
technology applied is completely different. An electromagnetic
tracking device,
miniBIRDTM fiom Ascension ~ e c h n o l o ~ ~ ' , dong with the
custom-designed sohare ,
Tremor Quantzfzer. is used to measure the fiequency, amplitude,
and area of displacement
of resting, postural, and intention tremor.
I Ascension Technology Corporation PO Box 527 Burlington,
Vermont, USA 05402 www.ascension- tech.com
DeveIopment of a Hand Tremor Quantincation Device for the
Measurement of Pathologid Tremor 2
-
Chapter 2 Background
Chapter 2 BACKGROUND
2.1 Parkinson's Disease
Parkinson's disease is by no means the only source of
pathological tremor, but it is a
major one. It is also the quest to quanti@ parkinsonian tremor
that has initiated and
driven this project. As a tesult, the system presented here has
been tailored to masure
three types of tremor commonly associated with Parkinson's
disease (resting, postural,
and intention) and as such it is helpful to have some general
background on the disease.
Parkinson's disease is a degenerative, neurological affliction
that usually occurs in later
life although occasionally may affect people as young as 30
years of age. One of the first
people to recognize parkïnsonian symptoms as a related
collective group was James
Parkinson. In his 18 17 An Essay on the Shaking Palsy he
describes the disease, which
now bears his name, as:
"hvoluatary tremulous motion, with lessened muscular power, in
parts not
in action and even when supported; with a propensity to bend the
trunk
fonvard, and to pass fiom w&g to a running pace: the senses
and
intellects being uninjured." p. 1
Pathologically, Parkinson's is characterized by the
deterioration of the nuclear masses in
the extrapyramidal system. It is a disorder of the basal nuclei
involving degeneration of
fibres fitom the sustantia nigra. Clinical symptoms include:
tremor of resting and
activated muscles, mask-like facies (Parkinson's facies),
slowing of voluntary
movements, festinating gai% peculiar posture, and muscular
weakness.
DeveIopment of a Hand T m o r Quantincation Device for the
Mea~urernent of Pathological Tremor 3
-
Chapter 2 Background
2.1.1 Symptoms of Parkinson's Disease
The symptoms of Parkinson's disease appear g r a d d y and
progress slowly. Tremor is
the most common and visible symptom of the disease. However,
other symptoms can be
more disabling. For example, hypokinesia (reduced movement) or a
feeling of
sluggishness may make it difficult for the penon to initiate
movement such as getting up
out of a chair or rnoving fiom a standing position to walking.
However, once the person
is in motion he/she is generaliy able to walk for a prolonged
penod of t h e . The gait is
often shunling and there is involuntary acceleration as the
penon seeks to readjust his or
her centre of gravity. Rigidity is another symptom and often
strikes in the muscles of the
arms, legs, and face as the tone of skeletd muscles increases.
#en this occurs in the
facial muscles it causes a charac:eristic 'rnask-like'
expression. Despite the fact that
Parkinson's disease stems fiom a degeneration of brain tissue it
does not adversely affect
mental cognitive capacity.
The symptoms of Parkinson's rnay hclude al1 types of tremor
(Fhdley 1996) but is
mostly characterized by four types (Findley et al. 198 1 ):
1. resting tremor
3. postural tremor
3. intention tremor
4. cogwheel phenornenon
2.1.1.1 Resting Tremor
Considered to be the hallmark symptom of Parkinson's disease,
resting tremor affects
about 75% of ali patients (Elble and Kolier, 1990 p.118). The
tremor is observed when
Devetopmwt of a Hand Tremor Quantification Device for the
Measuremeot of Pathologicd Tremor 4
-
Chapter 2 Back-ground
the muscles are not voluntarily activated and the body part is
supported agauist gravity.
For example, in a classic display of resting tremor, the patient
is seated with hisfher arm
resting on the legs and a distinct tremor begins to manifest
itself.
Resting tremor usually begins on one side of the body in an
upper limb before spreading
to the same-side leg d e r about 2 years (Findey 1996). The
tremor can remah on this
side of the body for a number of years before the other side is
affected (Findley 1996).
Manifestation is most common in the distal muscles of
extremities, such as the hands and
fingers, but rnay also occur in the arms, legs, lips, tongue or
jaw. Resting tremor rarely
effects the head and this distinguishes it fiom other disorders
such as essential tremorl
(Findley 1996).
h its early stages, resting tremor c m usually be inhibited by
voluntary activation of
afTected muscles. For example, patients may disguise resting
tremor in the hand by
performing a purposehl action such as touching the head or
picking up a pen. However,
this technique usually works o d y with mild tremor. As resting
tremor progresses, it
becomes more continuous and does not disappear with voluntary
muscle activation.
Resting tremor is most pronounced during periods of mental or
physical stress and also
seerns to increase in the upper extremities when the patient is
waking. In gened, tremor
is worsened by actions which distract the patient's
attention.'
' Essential tremor is an idiopathic, heterogeneous group of
movernent disordm or tremon. That is, the cause is unknown and this
tremor is not related to other pathological tremors such as that of
Parkinson's disease. This unqlained symptorn is not necessady a
firnction of disability, pathophysiology, or hmdity (FindIey 1996)
although 30-50% of patients have a history of essentid tlmor in
tbeir f d e s ( M e , 1995 p. 165). - The physician may induce a
high-stress state by asking the patient to perform a mentally
chaiienging task such as counting backwards from 100 by a prime
number.
Development of a Hand Tremor Quantif?caiion Device for the
Measurement of Pathologicai Tremor 5
-
Chapter 2 Background
A particular form of resting tremor termed the pill-rolhg tremor
is pathognomonic of
Parkinson's. A patient with this characteristic tremor can be
seen rhythmically extending
and flexing hisher wrist with a grasping movement of the k g e s
and a superimposed
rotational movement produced by rhythmic pronation-supination of
the forearm (Elble
and Kolier, 1990, p. 1 18).
Resting tremor has been characterized by diierent sources as
having a fiequency of:
4 to 7 Hz (Webster, 1968)
4 to 5.3 Hz (Findley et al., 198 1, Findley and Gresty, 1984, p.
295)
4 to 5 Hz (Capildeo and Findley, 1984, p.7)
4 to 6 Hz (Jankovic, 1987, p. 109)
3 to 5 Hz (Elble and Koller, 1990, p. 1 18)
4 to 5.5 Hz (Buckwell and Gresty, 1995, p. 148)
2.1.1.2 Postural Tremor
Postural tremor is also a very common symptom of Parkinson's
disease, usually occurrîng
in the upper extremities. This tremor has a higher fiequency
than resting tremor with
reported ranges of:
5.5 to 8 Hz (Findley and Gresty, 1984, p. 297)
5 to 8 Hz (Jankovic, 1987, p. 109)
5 to 12 Hz (Elble and Koller, 1990, p. 1 19)
6 to 8 Hz (Buckweil& Gresty, 1995, p. 148)
However, postural tremor usuaiiy has a smder amplitude (Findley
and Gresty 1984, p.
295) than resting tremor. Its fkquency makes postural tremor
dEcul t to distinguish
fkom essential trernor, which hm a range of 4 to 12 Hz (Elble
and Koller 1990 p. 61), and
Development of a Hand Trexuor Quantification Device for the
Measuremeat o f f athological Tremor 6
-
Chapter 3 Background
the 8- 12 Hz cornponent of nomial, always-present physiologie
tremor. Postural tremor is
a symptom in 60% of Parkinson's patients and if this is the
first manifestation of
Parkinson's disease without other significant symptoms it can
easily be misdiagnosed as
essential tremor (Elble and Koller 1990, p. 1 1 9). Elble and
Koiier ( 1 990, p. 1 19) report
than in 10 - 20% of Parkinson patients postural tremor is the
only form of trernor
exhibited in the course of illness.
2J.1.3 Intention Tremor
Intention tremor, sometimes called temiinal or cerebellar
tremor, occurs towards the end
of goal-directed movement. For example, in a 'touch-the-target'
test the finger begins to
oscillate with increasing amplitude as it approaches its target.
This type of tremor is not
exclusive to Parkinson's disease as it manifests as the result
of various cerebellar injuries.
It rnay even be diagnosed as a unique and rare inherited
disorder (McDowell 1971, p.
170). However, as a symptom of Parkinson's disease it may be
more disabling than rest
tremor because it is caused, as opposed to calmed, by muscle
activation. The tremor is
usually contùied to the han& and arms and has approximately
the sarne fiequency as rest
tremor (Webster, 1968), or 3-5 Hz (Jankovic and Fahn 1980).
2.1.1.4 Cogwheel Phenornenon
A rhythmicai, repetitive alteration in resistance during passive
movement of a limb about
a joint is temed the "cogwheel phenornenon". Cogwheeling has
been reported to
manifest itself in two distinct fiequencies: 6 Hz and the 8 to 9
Hz band (Findey et al.
198 1). Because the upper range is similar to the fkquency range
of postural tremor some
- - -
Development of a Hand Tremor Quantincation Device for the
Measurement of Pathologicai Tremor 7
-
Chapter 3 Background
researchers believe that the two tremor mechanisms are related
(Elble and Koler, 1990,
p. 119) while others daim a relationship to exaggerated
physiologic tremor (Webster,
1968). Although cogwheel tremor is often observed in Parkinson's
patients, it is not
exclusive to the disease and therefore cannot be used as the
definitive evidence for a
Parkinson diagnosis.
2.1.2 Treatmeut
Pathologically, parkinsonism (the symptom complex associated
with Parkinson's disease)
is the result of the stopping or slowing down of dopamine
activity in the corpus striatum
of the brain. This results fiom degeneration that occurs in the
dopaminergic nigrostriatal
pathway. Normally, the two opposing neurotransmitters, dopamine
and acetylcholine,
are balanced, but when dopamine is depleted, the overactivity of
acetylcholine produces
the symptoms of Parkinson's disease.
2.1.2.1 Anticholinergic Dnigs
For mild symptoms in patients with minor fùnctional impairment,
anticholinergic dmgs
such as trihexylphenidyl (Artane) or antihistamines with
anticholinergic properties, such
as diphenhydramine (Benadryl), c m be used effectively. These
types of agents block the
muscarinic effects of acetylcholine in the central nervous
system.
Cornmon side effects of anticholinergics include: dry mouth,
constipation, paralysis of
accommodation in the eye, and urinary retention. Less cornmon
but more severe
reactions may include memory loss and serious states of
confusion in elderly patients,
particularly those with dementia (Elble and Koller, 1 990: 1
27). These side effects usually
disappear within several days of treatment termination.
-- --
Development of a Hand Tremor Quantifidon Device for the
Measiwment of Paîhological Tremor 8
-
Chapter 2 Background
Most patients receive anticholinergic dmgs early in their
treatment, foilowed eventually
by levodopa or dopaminergic agonists. Cessation of
antichoiinergic treatrnent should be
done by tapering the dosage as opposed to rapid withdrawal
because this can result in
M e r exacerbation of symptoms, especially tremor (Elble and
Koller, 1990).
2.1.2.2 An tidepressants
Tricyclic antidepressants such as imipramine or amitriptyline cm
also be effective for the
treatment of mild Parkinson symptoms because they block the
re-uptake of dopamine
fiom the nerve synapses and also have anticholinergic effects.
However, it is not ciear if
these drugs improve motor signais directly or only have an
indirect effect through the
treatment of aaviety and depression.
2.1.2.3 Levodopa (L-dopa)
In patients with more severe symptoms, the augmentation of
dopamine levels in the brain
is required. This is often done with the dmg levodopa (L-dopa),
a breakthrough in the
treatment of Parkinson's disease. Levodopa crosses the
blood-brain barrier and is
converted to dopamine by decarboxylation, thus partially
correcting the dopaminergic
deficit within the striatum. Levodopa is the most effective
treatment for the entire range
of Parkinson symptoms. However, it does not always produce
consistent results,
sometimes entirely suppressing tremor in one patient but not
another without apparent
explanation. It has a particulariy bad track record for easing
resting tremor.
The decarboxylation reaction may also occw in penpheral tissues
causing adverse
reactions such as cardiac stimulation (tachycarida and
arrhythmias), confusion, nausea
and vomiting.
Development of a Hand Tremor Quantincation Device for the
Measurement of PathoIogical Tremor 9
-
Chapter 2 Background
2.1.2.4 Sinemet
A newer d m g than levodopa, called Sinemet, combines the
chemical carbidopa with
levodopa. Carbidopa works to inhibit production of dopamine
outside of the brain thus
lessening the side effects associated with the use of levodopa
alone. Sinemet
occasionally has only a limited span of effectiveness in some
patients and hence is
usuaiiy reserved for severe cases.
2.1.2.5 Dopaminergic Agonists
Dopaminergic agents such as bromocriptine, pergoiide, and
lisuride are also used to treat
Parkinson's disease. Although these drugs have not been shown to
be any more effective
than levodopa for resting tremor, occasionally a patient
responds dramatically to this type
of dmg.
Dopaminergic agonists are often used in combination with
levodopa and produce similar
side effects as those associated with levodopa.
2.1.2.6 Beta-Adrenergic Blocken
Many patients with Parkinson's disease experience postural
trernor that can be treated
with beta-adrenergic blockers such as propranolol or
nadolol.
2.1.2.7 Thalamotomy
Thaiamotomy is a stereotactic surgicd operation that destroys
specinc ceils in the brain
whose degeneration is Linked to Parkinson's disease.
-- pp
Development of a Hmd Tremor Quantification Device for the
Measurement of Pathologicai Tremor 1 O
-
Chapter 2 Background
Thalamotomy is usually perfomed unilaterally and produces a
reduction in resting
tremor in the contralateral extremities. The complications of
thalamotomy include
hemiparesis (partiai paralysis on one side), seizures, dysarthna
(imperfect articulation of
speech), dystonia (sustained muscle contractions), limb ataxia
(failure of muscuiar
coordination), confusion and hemidysesthesia (disorder of
sensation affecthg only one
side of the body). Bilateral thalamotomy has the increased risk
of postoperative
confusion and disorientation and also may result in dysarthria,
hypophonia (weak voice)
and dysphagia (difficulty swallowing) (Goldman and Kelly, 1 995
p. 554).
2.2 Classification of Tremor
Tremor is simpl y defïned as an involuntary , approximatel y
rhythmic, oscillatory
movement of a body part The amplitude of tremor is often so
srnaIl that it cm only be
measured with a highly responsive sensor. In other cases, such
as with Parkinson's
disease, tremor may be quite prominent.
Tremor can be divided into two very basic groups: physiological
and pathological.
Physiological tremor is a ubiquitous, asymptomatic (normal)
shaking that affects
everyone and results fiom the activity in individual motor
units. It is generally a low
amplitude (< 0.5 mm peak-to-peak, Elble et al. 1990), hi&
fiequency movement
(between 8 and 12 Hz, Findley 1996) that is barely visible to
the eye. People are rarely
aware of this tremor and it may only become noticeable when
tryhg to complete a fine
motor task such as threading a needle. In certain cases such as
fear or excitement, the
tremor might increase in amplitude to such a degree that it
interferes with simple actions
such as writing or holding a cup (Findley, 1996).
- -- -- -
Development of a Hand Tremor Quantincation Device for the
Measurement of Pathological Tremor 1 1
-
Chapter 2 Background
Pathological tremor, on the other hand, is that which arises as
a result of a disease or
disorder, usually of the central or peripheral nervous system.
Generally, if a tremor is
clinically visible and persistent it is consider pathological
(Findley, 1996). Tremor itself
is not a disease, but rather a symptoml. These tremors always
involve rhythmical
contractions of muscle groups (Findey, 1996 j which manifest in
periodic, that is, roughly
sinusoidai, movement about an axis.
2.2.1 E tiology
Tremor c m be organized or classified according to a number of
criteria. It has been
suggested that it would be useful to group tremor according to
etiology. However, it is
very difficult to associate a specific type of tremor with a
particular disease because many
tremors resulting fiom different diseases exhibit similar
characteristics. Furthemore, a
disease of the nervous system may result in more than one type
of tremor.
2.2.2 Amplitude, Waveform, Frequency
The rhyihmic nature of tremor suggests characterization by
amplitude, waveform or
fiequency. Amplitude, while important in tems of syrnptoms, does
not provide enough
unique information for categorization. Amplitude of tremor c m
be influenced by a wide
range of physiological, psychological, and environmental factors
and even under
controiled conditions displays considerable variability (Findley
1996). Tremor waveform
is also not very useful since there is no waveform that is
unique to a paaicular disease.
' Essentiai tremor is an exception. --
Development of a Hand Tremor Quantincation DeMce for the
Measurement of Pathotogicai T m o r 12
-
Chapter 2 Background
Frequency of tremor is its most stable parameter (Buckwell and
Gresty, 1995: 148) and is
often used for general classifications as can be seen in Table
2.1.
Table 2.1 Classification of common tremors by fiequency and
behavioural appearance Frequency Disease processAocus of lesion
Behavioural characteristics
1 2.5 -3 -5 Cerebellarhrainstem P o s ~ e t i c
Multiple sclerosis PostutaVkinetic Alcoholic degeneration
PostutaVkUietic Post-braumatic PosturaVkinetic
4 - 5 Parkinson's disease Rest Cerebellar disease
PosturaVkinetic Rubral Rest/posture/movement/kinetic Drug induced
Rest
5.5 - 7.5 Essential tremor PosturaVkinetic Clonus Parkinson's
disease Drug induced PosturaVkinetic
8 - 12 Enhanced-physiological tremor Postural/kinetic Drug
intoxications Essential tremor Cerebrocortical
Source: Fuidey 1996
However, although certain fiequencies are more typical of
individual diseases than
others, there is considerable overlap and that makes
classification by fiequency difficult.
2.2.3 Pharmacological Response
Another attempt to classify tremor is based on pharmacological
response. Some diseases
respond particularly well to specifk dnig therapies. For
example, Parkinson's disease
responds weli to dopaminergic dmgs and a response to smaii
amounts ofalcohol is faKly
indicative of essential tremor (Findley 1996). However, most
dmgs used to treat tremor
lack specinci~ (Findley 1996). AU dmgs with sedative action will
dampen tremor to
Developlnent of a Hand Tremor Quantincation Device for the
Measurement of Pathologicai Tremor 13
-
Chapter 2 Background
some degree. As a result, characterization by h g response is
not specific enough to be
particularly useful.
2.2.4 Appearance and Behavioural Characteristics
The lack of any other clear distinguishing feahire leads the way
for the most commonly-
used classification systern for tremor: a basic system that
categorizes tremor based on its
a p p e m c e and behavioural characteristics. The most cornmon
tremors are described
below. These definitions were drafted at the initiai meeting of
the Tremor Investigation
Group (TRIG) in Houston, Texas, in December 1990 as reported in
FUidley & Koller,
1995 (p. 1 - 2) and Findley, 1996.
2.2.4.1 Resting Tremor
This tremor is defhed as that which occurs when the muscles are
not voluntarily
activated and the body part is supported against gravity.
2.2.4.2 Action Tremor
This is tremor that occurs on the voluntary contraction of
muscles and includes postural,
kinetic, and isometric trernor.
2.2.4.3 Postural Tremor
Men voluntariIy maintaining position against gravity the tremor
that may result is
termed postural tremor. This type of tremor can be seen when the
arm is held in an
outstretched position. That is, the muscle must be activated in
order to keep the ann in
this position.
Development of a Hand Tremor Quantikation Device for the
Measurement of Pathologicai Tremor 14
-
Chapter 2 Background
2.2.4.4 Kinetic Tremor
This is the tremor that occurs during any sort of movement. Two
subsets of b e t i c
tremor include intention tremor and task-specific kinetic
tremor.
2.2.4.5 Intention Tremor
Also temed 'terminal tremor', intention tremor exhibits
progressive wonening towards
the end of a goal-directed movement such as touching a
target.
2.2.4.6 TaskSpecific or Occupational Kinetic Tremor
This tremor occurs or wonens during the carrying out of a highly
specific, skilled
movement. An example of this would be writing tremor.
2.2.4.7 Isometric Trernor
Isometric tremor is exhibited as a result of muscle contraction
against a @id, stationary
object.
2.2.4.8 Essential Tremor
This is an idiopathic, heterogeneous group of movement disorders
or tremors. That is,
the cause is unknown and this tremor is not related to other
pathological tremors such as
those of Parkinson's disease. This unexplained symptom is not
necessarily a fûnction of
disability, pathophysiology, or heredity. (Findley 1996)
2.2.4.9 Static Tremo r
Static tremor is a confushg t e m as it is used by different
researchers to denote either
resting or postural tremor. It WU not be used in this text.
Deveiopment of a Hand Tremor Quantification Device for the
Measurenient of Paîhologicai Tremor 15
-
Chapter 3 Background
1 Physiological 1 1 I
i 1 j Pathological 1 1
Figure 2.1 Tremor classification by appearance and behaviourai
characteristics
Development of a Hand Tremor Quantification Device for the
Measurement of Pathological Tremor 16
-
Chapter 3 Measurement of Tremor
Chapter 3
MEASUREMENT OF TREMOR
3.1 Review of Literature
Tremor lends itself well to measurement because of its rhythmic
and oscillatory
characteristics. An ideal tremor measuring device would be a
low-cost, accurate,
electronic device that quantifies tremor characteristics such as
fiequency and amplitude.
Several different electronic sensors are used for this purpose
but since none have been
perfected, subjective rating scaies are still widely used in
place of quantitative
measurements.
3.1.1 Subjective Assessrnent
3.1.1.1 Tremor Rating Scales
A number of tremor rating scaies have been developed and used
but none has become a
univenally accepted standard. The sensitivity of these scales is
insufficient to mesure
slight changes in tremor amplitude or fiequency. Nevertheless,
because of their
simplicity these tests continue to be widely used in the
clinical assessrnent of tremor
symptoms.
A few of the most common scales are listed in Table 3.1 - 3.3.
'
' Note tfiat one of the most popuiar rating scales for
Parkinson's disease, the Hoehn and Yahr Scale, îs not included
because it is not specific to the measurement of tremor but instead
rates the severity of the disease as a whole. This sale can be
found in Appendix A.
-- -
Development of a Hand Trernor Quantification Device for the
Measunment of PaîhoIogical Tremor 17
-
Chapter 3 Measurernent of Tremor
Table 3.1 Webster Scale for ~arkinsonian tremor 1 O = 1 No
detectable tremor found. 1
1 =
2 =
Table 3.2 Unified Ratine Scale for ~arkinsonian tremor' Version
2.0 - Decernber 1985
Less than 1 inch of peak-to-peak tremor movement observed in
lirnbs or head at rest or in either hand while waiking or during h
g e r to nose testing. Maximum tremor envelope fails to exceed 4
inches. Tremor is severe but not constant, and ~atient retains some
control of the hands.
3 =
- -- - - -- -
Tremor (as art of assessrnent of Activities of Daiiv Living)
Tremor envelope exceeds 4 inches and is constant. Tremor is
constant and severe. Patient cannot get free of tremor while awake
unless it is a pure cerebellar type. Writing and feeding are
impossible.
O = Absent 1 = Slight and infrequently present 2 = Moderate;
bothersome to patient 3 = Severe; interferes with many activities 4
= Marked; interferes with most activities
Tremor at Rest (as Motor Examination)
Source: Webster 1968
O = Absent 1 = SLight and aequently present 2 = Mild in
amplitude and persistent; or moderate in amplitude but only
intermittentiy present 3 = Moderate in amplitude and present
most of the t h e 4 = Marked in amplitude and present most of the
time
Action or Postural Tremor of Hands (as Motor Examination) O =
Absent 1 = Siïght; present with action 2 = Moderate in amplitude;
present during action 3 = Moderate in amplitude with posture
holding as well as action 4 = Marked amplitude; interferes with
feeding
Source: Koller 1 987 : 482-485 N e w versions of this rating
scde are termed: Unifïed Parkinson's Disease Rathg S d e
(LTPDRS)
- - - - - - -
~ e v e l o ~ m ~ of a Hand Tremor Quantification Device for the
Measurement of Pathological Tremor 18
-
Chapter 3 Measurement of Tremor
Table 3.3 Rating scale for essential tremor
1 Right Leît Hand: resting
postural kinetic intention
Finger (iso lated) Arm Les
Jaw
Voice
Tremor scale: O = absent: 1 = mild; 2 = moderate: 3 = severe
Head no-no yes-yes cornplex
Source: Elble and Koller 1990, p. 12
C h c a l assessrnent of tremor using these scales, before and
after treatment, can be used
to determine whether the therapy has been successful in calming
the tremor.
Other methods of measuring tremor include rating tasks such as
handwriting and drawing
Archimedes spirals'. Handwriting samples are ranked on the basis
of clarity and
smoothness and spirals are ranked according to tremor amplitude
and time taken to
accomplish the task. These evaiuation protocols s a e r fiom the
fact that they require the
subject to hold a pen in hiskier hand. Only tremor that is
superimposed on hand motion
during the act of writing or drawing can be measured. This
precludes the ability to
rneasure the most common tremon of Parkinson's disease: resting,
postural, and
intention.
m e r general testing protocols for measuring band tremor
include assessing the ability to
drink fiom a g las without spilling, timed h g e r tapping, and
pegboard tasks (Elble &
Koiler 1990, p. 1 1- 12). The difliculty with these action
tests, aside fiorn their
' Archimedes spiral:
pppppp
Development of a Hand Tremor Quantification Device for the M m e
n t of Pathologid Tremor 19
-
Chapter 3 Measurement of Tremor
subjectiveness, is that they may require excessive rnotor and
cognitive ninction. As a
result they may not be a pure measure of tremor. (Elble &
KolIer 1990, p. 1 1)
3.1.2 Objective Assessrnent
The diniculties inherent in subjective assessrnent of tremor
intensity have led researchers
to many attempts at evaluating tremor in some objective,
quantifiably measurable
manner.
3.1.2.1 Tambour
The earliest attempts to measure tremor made use of a tambour
applied to a Limb. The
limb movement resulted in displacement of this receiving tambour
which was affixed to
the limb under investigation. The receiving tambour was commonly
attached via mbber
to a second recording tambour whose lever was connected to a
rotating cim. A pen set
against the d m made a record of the limb movement. Several
early researchers
describe the set-up of their experiments used to investigate
muscle contraction and tremor
(Horsley & Schafer 1886, Wolfenden & Williams 1888,
Eshner 1897).
This early method was satisfactory in many respects,
particularly in its tirne. However,
the sensitivity and accuracy of a variety of newer technologies
greatly surpasses that of
the ancient tambour.
3.1.2.2 Optical Methods
Other early researchers opticdy magnined the displacement of
light caused by iimb or
finger movement. This Somat ion was transduced by a
photo-electric cell and
Development of a Hand Tremor Quantincation Device for the
Measurement of Pathological Tremor 20
-
Chapter 3 Mesurement of Tremor
amplined. The output of this was then fed to a moving-coii pen
or oscilloscope (Jasper
and Andrew 1938, Graham 1945, Cooper et al. 1957).
Because these methods could only record srnall movements within
a lirnited space
restriction they were only used for poshiral and resthg
tremor.
By far, the most widely used method of quantifjmg hand tremor is
accelerometry. Many
researchers, dathg back over a hundred years, have used this
rnethodology to measure a
variety of body tremors (Boshes 1966, McAllister et al. 1985,
Ghika 1993, Timmer et ai.
1996). Uniaxial or triaxial accelerometers are widely available.
Tremor is rarely, if ever,
a unidirectional movement (Elble & Koller, 1990) and since
uniaxial accelerometers only
measure tremor in one dimension they have serious limitations
Triaxial accelerometers
are now becoming more popular as their cost and size decrease.
Despite this, the
majority of researchers in this particular field have used
uniaxial accelerometers with
only a few using multidimensional sensors (Jankovic and Frost,
198 1, Salzer, 1972).
With respect to measuring tremor, miniature accelerometers have
the advantage of being
quite srnall and iight-weight so they do not have a dampening
affect on the trernor. In
addition, the sampling rates offered by some accelerometers are
far greater than one
would even need to measure motion up to 20 Hz. Furthemore,
accelerometers are much
more sensitive to high fiequency vibrations than displacement
tramducers (Scholz &
Bacher, 1995, p. 295, Gresty and Findey, 1984 p. 17).
Development of a Hand Tremor Quantification Device for the
Measurement of Pathological Tremor 21
-
Chapter 3 Measurement of Tremor
These attributes suppori the use of accelerometers for
determining the fiequency
component of tremor. However, there are a number of chical tests
that require
information about the position of the hand. For example, a
patient may be asked to trace
a certain shape with a sensor on hisher hger . Amount of
deviation fiom the trace shape
indicates severity of tremor. Positional information is oot
easily obtained fiom
acceierometer data. The amplitude of tremor cannot be determined
and it has been noted
that the tremor a m p h d e denved fiom acceleration does not
correlate well with the
functional disability experienced by the patient (Spyers-Ashby,
1997, p. 36). On the
other hand, absolute amplitude of tremor and the impairment and
disability it can cause
are of interest to the patient and clinician (Biickwell and
Gresty, 1995, p. 148).
3.1.2.4 Digitizing Ta blets
Digitizing tablets have also been used by researchers (Elble et
al. 1990, Pullman 1998) to
quanta trernor. In this approach, a digitking tablet is normally
linked to a computer via
an RS232 serial connection and specimens of handwriting or
Archimedes spi& are
analyzed. What was on paper a subjective judgement of clarity
and smoothness, can now
be quantified in the extracted amplitude and fiequency
information of the superimposed
tremor. Commercial digitizen can have resolutions that range fhm
78.7 to 787 ünes per
centimetre (0.127 mm to 0.0 12 mm), accuracies of 0.25 mm, and
data rates of 200 points
per second (Elble et al., 1990). These tablets are a good
alternative to subjective rating
scales because they provide some quantitative rneasure of tremor
at a relatively low cost'.
However, they can only be used to measure the tremor that
accompanies tasks such as
' As of this writing a 4" x 5" Wacom Intuos digitking tablet was
priced at $200 US [Accuracy: iO.25 mm, Resolution: 200 Lines per
mm, Max data rate: 200 points per second)
Development of a Hand Tremor Quantincation Device for the
Measurement of Pathologicai Tremor 22
-
Chapter 3 Measurement of Tremor
writing or drawing with a pen in hand. Resting tremor, which is
the most prevalent
symptom of Parkinson's, cannot be measured in this manner.
3.1.2.5 Optoelectric Imaging Systems
The gold standard of three-dimensional position tracking devices
are optoelectric irnaging
systems, such as the OPTOR4K (Northem ~igital)', which use CCD c
m n s to
determine the location of h f k e d markers (LEDs). The OPTOTRAK
in particular
provides an accuracy of 0.1 mm and a sampiing rate of up to 3500
Hz. Three CCD
cameras with Light sensitive pixels measure the amount of energy
striking each pixel to
determine the position of the light source. Although the
accuracy of these systems is
enviable, their price tag makes them out of reach for the
individual clinician. The cost
range of such systems is in the tens of thousands of dollars. In
addition, the equipment is
fairly large (the CCD unit stands about a metre high) and a
direct line-of-sight must be
maintained between the idiared LED and the cameras.
3.1.2.6 Surface Electromyography
Surface electromyography @MG) quantifies the muscular activity
causing tremor. As
such, it is not a direct measure of tremor itself and is most
ofien combined with another
measurement device (e.g. accelerometer) in order to investigate
the way in which motor
unit activity and tremor are related.
' Nonhem Digital hc 103 Randaii Dr. Waterloo, Ontario, Canada
N2V lC5 www.ndigital.com Development of a Hand Tremor
Quantification Device for the Measurement of Pathological Tremor
23
-
Chapter 3 Measurement of Tremor
3.2 Tremor Characteristics
In order to choose an appropriate measurement device it is
important to understand the
characteristics of what is king measured. For example, when
measuring kinetic tremor a
device is required that is sensitive to both tremor and
voluntary motion. However, when
mesuring resthg tremor, voluntary movement is not important. The
best recording
device is that which has signal-response characteristics that
are similar to the signal
properties of the tremor.
3.2.1 Waveform
Tremor lends itself well to measurement because it is a roughly
penodic, sinusoicial
signal. Although the system presented here investigates motion
in three dimensions it is
instrumental to consider the one-dimensional characteristics of
tremor. In its shplest
form, a tremor can be expressed as:
d = ~ s i n ( 0 t ) [Eqn. 3.11
Where d = positional displacement; A = peak amplitude; t = tirne
and o = firequency of
oscillation (rads). Although tremor is never perfectly
shusoidai, the Fourier theorem
states that the signal c m be analyzed into a series of sine
waves of appropriate amplitudes
and fiequencies.
The first derivative of equation 3.1 provides the velocity of
movement:
v =Aocost [Eqn. 3-21
Note that the peak velocity (when cos t = 1) is simply the peak
amplitude of displacement
multiplied by the fkquency of oscillation in radians per
second.
Development of a Hand Tremor QuantiiIcation Device for the
Measurement of PathoIogical Tremor 24
-
Chapter 3 Measurement of Tremor
The second derivative of equation 3.1, or the Est derivative of
equation 3.2, provides the
acceleration of movement:
[Eqn. 3.31
3.2.2 Frequency
Although there is some debate about whether it is possible to
distinguish between
physiological and pathological tremor based on fiequency alone,
there is no doubt that
fiequency is one of the most important measures of tremor
(Deusch1 et al. 1996).
Although the fkequency ranges for pathological and physiological
tremor overlap, some
conclusions can be inferred fkom a given fiequency measurement.
Ln 1996, Deuschl et al.
measured the fkequency distribution of hand tremor for normal
controls, patients with
Parkinson's, and patients with essential tremor. Their work
showed that physiological
tremor clearly fits between 6 and 11 Hz, whereas pathological
tremor (essential and
parkinsonian) have significantly lower peak fiequencies between
4 and 10 Hz.
Regarding the 6 Hz breakpoint, almost none of the normal
subjects exhibited a tremor of
under 6 Hz; 25% of patients with essential tremor exhibited a
peak fiequency of 6 Hz or
under; and 65% of Parkinson's patients had a tremor equal to or
lower than 6 Kz. The
conclusion fkom this is that although fiequency may not be
enough to completely
distinguish between pathological and physiological tremor it
plays a signincant role and
is therefore an important panuneter of tremor.
The issue in terms of assessing the benefit of treatment (i.e.
phannacologicd or surgical)
is whether it subdues the tremor and by how much. In this
instance it is not a matter of
diagnosing an illness or type of tremor. Thus the question
related to kquency is: Has
- -
Development of a Hand Trenior Quantification Device for the
Measunment of Pathological Tremor 25
-
Chapter 3 Measurement of Tremor
the fiequency of the tremor become more 'normal'? That is, has
the Wemor nequency
moved up into the 6-1 1 Hz band?
Although tremor fiequency does fluctuate within the penod of a
day (see Section 3.2.4
below) it does not appear to fluctuate over moderate periods of
t h e . That is, al1 things
being equai ie.g. level of stress, rime of day, amount of
cafi?eine consumedj a person's
tremor does not change sigmfïcantly ftom day to day. Data taken
over a 3-&y period
have shown that tremor fiequency is stable in patients with
Parkinson's disease and over a
penod of 5 years there is only a slight decrease in tremor
fiequency of less than 2 Hz
(Deuschl et al. 1996). As a result, a change in fiequency of
tremor after surgical or
pharrnacological treatment wouid Likely be an indication of a
treatmeot's efficacy.
3.2.3 Amplitude
The amplitude of tremor is not appreciably helpful in
differentiating between various
types of pathological tremor but can distinguish between
pathological and physiological
tremor. Patients with pathologicd tremor generally have much
higher tremor amplitudes
than nomal subjects. Furthemore, the absolute amplitude of
tremor and the impairment
that accompanies it can be a very important factor for both
clinician and patient
(BuckweLl and Gresty, 1995, p. 118).
3.2.4 Long-Term Tremor Measurement
Most of the studies conducted to measure pathological tremor
have concentrated on
tremor exhibited within a very short tirne frame, usually a few
minutes. However, it
pp -- - - - - - - - -- -
Developrnent of a Hand Tremor Quantification Device for the
Measurement of Pathologicai Tremor 26
-
Chapter 3 Measmement of Tremor
should be noted that trernor amplitude does undergo diunial
fluctuations (Elble and
Koiler, 1990, p. 24).
Without takllig into consideration tremor fluctuations over the
course of a day, it is
àiflicult to assess the true disabling nature of the tremor.
Mild tremor that is consistent
over the course of the day cm be as debilitating as severe
tremor that oniy manifests itself
occasionally. In order to account for these variables, some
researchers require subjects to
peIform trernot enhancing tasks (e.g. mental arithmetic,
"Stroop" test1) in order to record
the maximum amplitude tremor possible (Elble and Koller, 1
990).
ûther researchers have attempted to measure tremor over extended
periods of time in
order to more accurately assess disability (Smeja et al. 1999,
Spieker et al. 1998).
Because d ima l fluctuations are so difficult to measure, their
effect should be taken into
account by experimental control: taking measurements for the
same length of time and at
the same time of day.
3.3 Selection of a Measurement Device
The eventuai choice of Ascension's mlliiBIRDm electromagnetic
tracking system for the
application of hand tremor quantification came only after the
evaluation of a number of
different systems. The technologies considered are bnefly
descnbed below.
' h the Stroop test the patient is presented with words - names
of common colours (mi, yeilow, green, blue). Each word is printed
in a coloured ink that ciiffers tiom the word-name. For exampIe,
the word "Yellow" might be printed in blue ink The patient is asked
to name the colour of the inks in which each word is wina.. This
requires an inîeUectu;al effort because reading is a more
over-Iearned response than colour naming and therefore the patient
has to inhibit the dominant tendency to say the word (Gresty et al.
1984, p. 323).
- - - - -- -
Developrnent of a Hand Tremor Quantification Device for the
Measmernent of Pathologiçal Tremor 27
-
Chapter 3 Measurement of Tremor
Interest in revisiting the idea of developing a hand tremor
quantification system came
with the accidental discovery of a wireless, 3D tracbg
RingMouse. See Figure 3.1.
Figure 3.1 : Pegasus RingMouse
This product is made by Pegasus Technologies ~ t d ' . (Holon,
Israel) which markets the
device as a wireless joystick (the RingMouse is mounted in a
handle) for use in computer
games such as Doomm or QuakeTM. It generates x, y, z coordinates
in real time using
bot , &ed and ultrasonic signals. Two ASICs (Application
Specific Integated
Circuit) are incorporated into a receiving unit that is mounted
on the computer monitor.
The first ASIC includes signai processing software aigorithms
and performs reai-tirne
' Pegasus Technologies Ltb MerkaPm 2000. 5 Hazoref St Hoion, h e
l , 58856,. www.pegatecbcom Development of a Hand Trernor
Quantification Device for the Measurement of Paîhological Tremor
28
-
Chapter 3 Measmement of Tremor
calculations while the other contains condensed signal
processing hardware. Ultmonic
signals are transmitted fiom the wireless, mobile RingMouse to
the receiving mit. Using
DTO A (DifTerential Times Of Amval) technology, the receiving
unit calculates the
location of the RingMouse by use of tnangulation calculations.
The buttons on the
RingMouse transmit information to the receivuig imit via
infriired signals. In addition,
because the system's power supply is less than 10 mA, power can
be drawn directly fiorn
the cornputer's serial port and no extemal power supply is
required.
The RingMouse appeared to be an excellent option because of its
low cost (475 US),
light weight, and accuracy. However, it had a maximum sampling
rate of 50Hz
according to the manufacturer's specifications. Furthemore,
practical experimentation
was only able to c o n h a sampling rate of 20 to 30 Hz. This is
inadequate for
measuring signals up to 20 Hz. Initial communication with
Pegasus regarding a
customized version of the RingMouse with a higher sampling rate
appeared promising.
However, the pnce for the design alterations as quoted by
Pegasus was prohibitive and
the idea was abandoned.
Mer the RingMouse technology was deemed unsuitable, an
investigation was launched
into alternative measurement systerns. As mentioned earlier, the
majority of the work
done in haod tremor quantification has used accelerometrîc
technology. An extensive
literature review was undertaken concurrently with a survey of
miniature, triaxial
accelerometers on the market. Two Merent accelerometea were
brought into the lab for
Development of a Hand Tremor Quantification Device for the
Measurement of Pathologicai Tremor 29
-
Chapter 3 Measmement of Tremor
assessment: Brüel & Kjær's model 4507 and Entran's model
EGAX-F. The Entran
accelerometer is shown in Figure 3.2.
Figure 3.2: Close-up of Entran EGAX-F miaxial accelerometer
Both accelerometers were uniaxial but gave an idea of the
accelerorneters' response. In
the case of the Entran EGAX-F, two uniaxial accelerometers were
mounted perpendicular
to each other to obtain information about the motion in both the
x and y directions. The
accelerometers' response was excellent and the decision appeared
only to hinge on the
choice of a suitable Light-weight, miniature, triaxial
accelerometer that met al1
requirements. However, in the end this approach was also
abandoned. This decision was
dtimately made based on the accelerometer's data output.
Although accelerometers
- - --
Development of a Hand Tremor Quantification Device for the
Measurement of Pathologicd Tremor 30
-
Chapter 3 Measurement of Tremor
provide excellent and accurate acceleration data, which is
sufncient to obtain tremor
fiequency, the amplitude provided is that of acceleration.
Amplitude of acceleration is
not a usehi parameter in measuring tremor. Amplitude of tremor,
a required parameter,
c m be theoretically, but not practicaiiy integrated fiom
acceleration data.
3.3.3 Electromagnetic Tracking
Electromagnetic tracking technology proved to be the optimal
solution to the hand tremor
measurement dilemma. The positional information of a small
receiver unit with respect
to a magnetic coi1 transrnitter is relayed to an electronics
system which is typically
comected to the cornputer via a s e r d port (RS232 comection).
Two electromagnetic
tracking systems were evaiuated in this case. The first was the
miniBIRD~1 by
Ascension ~ e c h n o l o ~ ~ ' (Figure 3.3) and the second was
the FASTRAKm fiom
Poihemus hc2. (Figure 3.4).
The two systems use virtually identical technology and are very
similar in physical
appearance. See Figure 3 S.
' See Appendk B for complete specifications. ' Polhemus
hcorporated. 40 H d e s Drive P.O. Box 560 Colchester, VT, USA
05446 www.poihemus.com
- - . -. . - -
Development of a Hand Tremor Quantification Device for the
Measurement of Pathological Tremor 3 1
-
Chapter 3 Measurerneut of Tremor
Figure 3.3 rniniBRDm fiom Ascension Technology
Figure 3.4 FASTRAKm fiom Poihemus Inc.
Development of a Hand Tremor Quantincation Device for the
Mea~ufement of Pathological Tremor 32
-
Figure 3.5 : miniBIRDM venus FASTRAKTM
The main difference between the two systems is that the FASTRAKW
is an AC system
whereas the miniBIRDm uses DC technology. A complete cornparison
c m be found in
Appendix C. The critical differences between the two systems are
sensor sue, sampling
Eequency, susceptibility to metdlic distortions, and cost.
The miniBIRD'sm 18mm x 8mm x 8mm receiver is smaller and lighter
than that of the
FASTRAKT One of the major considerations in measuring tremor is
that the sensor
size and weight not dampen the actual tremor. As such, it is
important that the sensor be
as s m d and iight as possible. In addition, Ascension
Technology is currently developing
an even smaller sensor (IOmm x 5mm x 5mm) in their miniBIRD~
model500 which
will be implemented into this system when in cornes on the
market in the fd of 2000.
See Figure 3.6.
Development of a Hand Tremor Quantincation Device for the
Measuremmt of Pathologicd Tremor 33
-
Chapter 3 Measurement of Tremor
Figure 3.6: miniBIRDm sensor: model 800 versus model 500
Another technical consideration was the miniBIRD's maximum
sampling rate of 144 Hz
which was supenor to the FASTRAK'sTM 120 Hz. Also, the DC
magnetic technology
used in the rniniBLRDfM is reported to be Iess susceptible to
metallic distortion than the
AC technology used in the FASTRAKTM (Milne et al. 1996). The
final deciding factor
was price. At a cost of over one and a hdf times that of the
miniBIRDTM, ihe
FASTRAKm was not priced competitively. Its higher pnce likely
has to do with the four
channels that it provides versus the single channei that the
miniBIRDfM offers. If four
channels were required it would defmitely be more cost effective
to purchase a single
FASTRAKTM versus four single-channel miniBIRDsTM. However, this
application only
requires one measurement channel and this is provided by the
miniBIRDThf at a lower
cost.
Development of a Hand Tremor Quantification Device for the
Measurement of Pathological Tremor 34
-
Chapter 4 Methodology
Chapter 4
METHODOLOGY
4.1 Description of Technology : rniniBIRD-
The tracking device used to measure tremor motion is the
miniBRDm Model 800 fiom
Ascension Technology. The rniniBLRDTM is based on the same
technology as
Ascension's main product, the Flock of BudsTM (FOB), except that
with sensor
dimensions of 18- x 8 mm x 8mm it is the smallest probe that
Ascension makes' and
is specifically designed for biomedical applications.
The miniBIRDm memures the real-time position and orientation (6
degrees of fieedom)
of the receiver (sensor) with respect to a transrnitter using
pulsed direct curent magnetic
fields. It is made up of four main components: the transmitter,
receiver, electricai unit,
and power supply. See Appendk B for complete specifications. The
transmitter contains
three coils mounted orthogondly about a cubic core. Electrical
circuitry within the
transmitter sends puises of direct current ihrough each of the
three tnuismitter coils in
tum and this generates an elecûomagnetic field. Each receiver
also houes three
orthogonal coils which measure the respective components of the
magnetic field in which
the receiver unit is placed.
Several researchers have tested Ascension's FIock of BirdsTM and
concluded that it is
appropriate for various biomechanicai measurements (Mihe et al.
1996, Bottlang et al.
1998, Mesken et al. 1999, Bottlang et al. 2000).
Ascension Tecimology is ctarently working on the miniBIRDm Model
500 which wül have sensor dimensions of 10 mm x 5 mm x Smm (See
Section 3 3 3 )
Development of a Hand Tremor Quantification Device for the
Measurement of Paîhological remo or 35
-
Mihe et al. (1996) conducted an investigation into the accuracy
of the Flock of BirdsTM
based on its optimal operating range and the degree to which
metal interfered with the
electromagnetic signai. Their study concluded that the optimal
operation zone (that
which exhibited the smaliest positional error: < 2%) was
within a trammitter-to-receiver
separation range of 22.5 - 64.0 cm. In addition, for small step
positional increments (<
2.5 cm) the device was sensitive enough to obtain a resolution
of 0.25 mm. Angular
motion could be resolved to O. 1". These results were better or
equai to the manufacturer's
daims of accuracies of 2.5 mm RMS and 0.5" RMS averaged over the
translationai range
and resolutions of 0.75 mm and 0.10" at 30.5 cm.' Furthemore, it
was shown that only
rniid steel (a cylindx-ical piece 12 mm in diameter and 125 mm
in length) exhibited
significant intefierence on the data (other metals tested were
titanium, stainless steel,
cobalt chrome, and aluminum), particulariy when placed next to
the receiver (at 6 mm
from the edge). Overall, the device was deemed to be a useful
tool for a variety of
muscuioskeletal research investigations.
Bull et al. (1997) repeated Milne et d.'s work co-g the optimal
operating range but
claimed the positional accuracy to be an order of magnitude
better at large step sizes.
Their explanation was the fact that the Flock of Birdsm
maintains the strength of the
magnetic field at the receiver (sensor) by stepping up the
trammitter power as the
receiver is moved m e r away fkom the transmitter. Transients
during this process
adversely affect the system. However, by using a power supply
with a higher ratiq than
' Note tha< the latest generation of b o t . the FOBW and the
miniBIRDRL daim accuracies of 1.78mm 1 O.SO RMS and resolutions of
0.5 mm / 0.1" at 10.5 an
Development of a Hand Tremor Quantification Device for the
Measurement of Pathoiogicai Tremor 36
-
Chapter 3 Methodology
that supplied by the vendor, Bull et al. were able to obtain
smaller enors at the extreme of
transmitier to receiver separations.
Bottlang et al. (1 999) performed similar venfication tests with
comparable results. They
determined that the main source of data distortion was randorn
noise and not systernatic
error ( e g distortion &om ferromagnetic objects,
calibration erron of the receiver coils).
ûther researchers, namely Meskers et al. (1999) and Bottlang et
al. (2000), used the
Flock of B i r d s T M for upper body kinematics (shoulder and
elbow motion respectively)
and concluded that it was a useful tool.
4.2 Serial Interface
A software handshaking RS232 cable is used to connect the
miniBIRDm to the host
computer which captures the data acquired by the miniBirdTM.
Only pins 2,3 , and 5 of
the 9-pin interface connector are required for
communication.
Pin RS232 Siaa l Direction - Description 2 Receive Data BUd to
Host Serial data output fiom the BW to the host 3 Transmit Data Hoa
to Bird Serial data output fiom host to Bird 5 Signal Ground Signal
Reference
4.3 Hardware Configuration
The host computer used in this setup has a Pentium II', 400 MHz
processor with 64 MB
RAM, and the Microsoft Windows TM 952 operathg systern.
' Compaq Compter wwwcompaq.com ' Microsofi
www.mic.rosofi.com
Development of a Hand Tremor Quantincation Device for the
Meamment of Pathologicai Tremor 37
-
Chapter 4 Methodology
The miniBIRD'sTM dipswitchs are set to:
Dip Switch Position:
- - -- 1
Sets Baud aie to 19200 Sets address to "stand-alonel' mode
"Fly"
1 12 on 1 o n
The "stand-alone" mode is used because there is oniy one m i n i
B I R D T M in this
configuration (multiple miniBIRDsTM are required if more than
one sensor is used to
track motion). The "Fly" mode (dipswitch 8 of€) sets the
miniBIRDm to a state ready to
begin acquiring data. This is in contrast to the "Test"
(dipswitch 8 on) mode that is used
to run specific manufacturer-designed tests on the
mullBkdTM.
4.4 Software Description
The software T m o r Quczntifer was written by the author in
LabVEWTM 5.1 (National
instruments, Austin, exa as') and custom designed to interface
with the miniBIRD% It
has three display modes:
I ) Acquired Data (data acquired using the rniniBIRDRn)
2) Sample Data (previously recorded data)
3) Test Data (for illustrative and testing purposes)
Figure 4.1 is a shot of the main screen of Tremor
Quantzifer.
3 off
' Nationai instruments Corporation 1 1500 N Mopac Ekpwy A-
Texas, USA 78759-3504 www.ni.com
Development of a Hand Tremor Quantification Device for the
Measurement of Pathological Tremor 38
4 off
5 off
8 off
6 off
7 off
-
Chapter 4 Methodology
Figure 4.1 : Screen shot of Tremor Quantifier
Developrnent of a Hand Tremor Quantification Device for the
Measrwment of Pathological Tremor 39
-
The prLnary mode is the acquisition and display of tremor data
as obtained fiom the
miniBIRDfM. Upon depressing the 'Acquire Data' button the
program:
1. Reads data fiom miniBIRDfM.
2. Decodes data to determine x, y, z coordinates of
position.
The pro= - executes the remainine sequential actions regardless
of the type of data
mode chosen:
3. Displays motion in each axis (x, y, z) with 3 options:
1) Raw data
2) Mean subtracted fiom data
3) Data passed through hi&-pass filter
4. Displays power spectrum for each mis.
5. Estirnates dominant fiequency in each a i s .
6. Calculates Deviation From Mean @FM) for each data point (x,
y, z).
7. Detemiines peak and average amphde of tremor.
8. Determines area uademeath DFM plot per second.
9. Displays Signed Deviation From Mean (SDFM) c w e @FM cuve
with direction
taken into account).
10. Displays power spectrurn of Signed Deviation From Mean
(SDFM) data.
1 1. Estimates the dominant fkequency for overall tremor.
12. Provides option to store tremor information to file.
13. Provides option to export information to spreadsheet
nle.
14. Provides option to load previously-stored data.
15. Provides option to print out &ta andysis dong with
graphs.
This sequence of events is described in m e r detail below.
Development of a Hand Tremor Quanrification Device for the
Mea~ufement of Pathological Tremor 40
-
1. Readine Data From miniBIRD-
The Tremor @anhifier executes the foilowing steps to read the
data fiom the
rniniBIRûm:
1) Uutializes the serial port.
1 S t o ~ bits: 1 1 1
Port number: Baud rate: Data bits:
2) Writes to the serial port to set sampling rate.
Sends CHANGE VALUE command to miniRIRDm to set BIRD
MEASUREMENT
RATE. The CHANGE VALUE comrnand must be issued to the
minü3IRD'fM in the
followiag sequence:
O (Corn !) 19 200 8
1 BYTE # 1 1 t l 1 Command Bvte 1
The default rneasurement rate for the mllüBIRDm is 103 Hz. The
steps to set the
rneasurement rate to 144 Hz are detailed below. This general
procedure can be followed
to set the sampling rate to any other acceptable value (20 - 144
Hz). The program sends
four bytes of information to the miniBIRDfM:
Deveiopment of a Hand Tremor Quantiacation Device for the
Mea~ucement of Pathoiogical Tremor 41
Byte # 1 2 3
Hex Value 50 07 00
Command Byte PARAMETERnumber PARAMETERdata LSByte
4 1 90 PARAMETERdata MSBvte " PARAMETERvalue
-
Chapter 4 Methodology
The CHANGE VALUE command is:
The CHANGE VALUE command allows for the changing of the m i n i
B I R D T M system
parameter defined by the P.4,rtLWERnumber byte md the
PAwIETLRvdue bytes
Command Byte
sent with the command:
1 Command Data 1 PARAMETERnumber 1 PARAMETERdata 1
ASCII P
The PARAMETERnumber for BIRD MEASUREMENT RATE is 7.
To CHANGE the BIRD MEASUREMENT RATE, the program sends the
miniBIRDm
one word of PARAMETERdata correspondhg to: (measurement rate) x
256.
In order to set the BIRD MEASUREMENT RATE to 144 Hz
consider:
HEX 50
144 x 256 = 36 864 (Decimal) = 9000 (Hexadecimal)
Thus the Hex version of the PARAMETERvalue is 9000. That is:
DECIMAL 80
MS Byte LS Byte 90 O0
BINARY 0101 O000
X LS Byte MS Byte
Therefore, the Hex command sent to the miniBIRDm to change the
BIRD
MEASUREMENT RATE to 144 Hz is 50 07 00 90.
DeveIopment of a Hand Tremor Quantifidon Device for the M-ent of
Pathologicai Tremor 42
-
Chapter 4 Methodology
CHANGE VALUE BIRD LS Byte of 9000 MS Byte of 9000
MEASUREMENT
RATE
3) Writes data acquisition command to serial port.
The Tremor Qumtz~er program sends the command IV@' to the
miniBIRD~. ' Ibis
indicates a request to obtain POSITION information (x, y, z
coordinates) in STREAM
mode.
J) Reads fiom serial port, 6 bytes at a time (each coordinate
[x, y, z] has two bytes of information associated with it).
The program captures data for the time penod specified by the
user. If the user chooses
to capture 5 seconds of data then the program will capture 720
data samples:
capture period [ s ] x sampling rate = No. of sarnples
samples 5 ~ x 1 4 4 = 720 samples
S
Since each sample contains 6 bytes of data this means that 4320
bytes of data are
collected:
bytes 720 samples x 6 = 4320 bytes samples
Development of a Hand Tremor Quantification Device for the
Measmement of Pathological Tremor 43
-
2. Decoding data to determine x,va coordinates
The data read fkom the miniBRDm comes in the f o m of decimal
nurnbers and is fed
into an amy. The information is arranged in the may in the
following order:
Where Point O, Po = (h, y,, G) point 1, Pl =(xi, Yi, 21)
Although this is the order that the information comes in, it
does not always begin with the
LS byte of an x coordinate. In order to identify the beginning
of a sequence of data
representing a single point, the LS byte of every x coordinate
is flagged with a leading '1'.
In order to furnish the LS byte of every x coordinate with this
leading '1' the data is coded
LS Byte %
in a paaicular way. Therefore, the byte with the leading one
must be f o n d and then
the bytes must be decoded. in order to do this the Tremor
@unifier:
- - - * *
101 Pl Pl 131 141 Pl VI VI 181
MS Byte &
LS Byte Y0
1) converts the data into binary values
2) runs a search for the leading '1' and splits the array at
this point, discarding
MS Byte Y0
every byte to the left of the byte with the leading '1' (note
that at maximum
there can only be 5 bytes in the discarded array)
Mer catchhg the Ieading '1' the Tremor Quunt@er translates the
bytes into actual x, y, z
coordinates by the foIlowing method:
LS Byte 4
3) shifts each LS byte Ief3 one bit
J) combines each MS byte/LS byte pair into data words (1 word =
2 bytes)
MS Byte 2,
5) shifts each word left one more bit
- * - - -
LS Byte ; : Yi : ...{
LS Byte XI
Deveiopment of a Hand Tremor Quantifkation Device for the
Measmement of Paîhological Ttemor 44
MS Byte XI
-
Chapter 4 Methodology -- - -- - --
Then the program:
6) multiplies the decimal e q d e n t of the binary number by
the position
36 constant - to obtain coordinates in inches
32768
7) multiplies the coordinate values by 2.54 to obtain values in
centimetres
An example is used to illustrate this procedure:
The data that cornes fiom the bird is put into an array. A
sample array rnay be as follows:
The value in position [l] has the leading '1' so this is the LS
byte of the x coordinate.
Everything to the left of this is striped fiom the array. Now
the array is:
LS byte MS byte g LS byte y. MS byte y. LS byte 2. MS byte z,
----- - LS byte --a----, xi 1 1 1 1 1 1000 ~00000111 10011 1001
10110 1 1 1 1 ~01100001 10111 1101 1 *-.*,---*.**--l ..*
PI [ I I PI i31 PI 153 FI
Al1 the LS bytes are shifted lefi by one bit:
The MS byte/ LS byte pairs are combined into a data word:
Development of a Hand T m o r Quantification Device for the
Measirrement of Pathologicai Trernor 45
-
Chapter 4 Methodology
Values over 16 384 are in two's complement, indicating negative
values. Thus, taking the
two's complement of those values:
Then, each word is shifted to the left one more bit:
x, YO
1 1 B inarv 1 Shifi Left One Bit 1 Decimal 1
36 Each value is multiplied by the position factor - to obtain
coordinates in inches and 32768
Binary O000 11 11 11 10 O000 0110111101110010
multiplied again by 2.54 to obtain coordinates in
centimetres:
Two's Complement
10010000~0001110
2*54 cm 1 y. = - 9.3 1 incher x 2S4cm 1 z, = - 1 .?7incher x
2.54 cm xo = 4.46 Niches x inch inch inch
Decimal 2 032 -4 238
36 .r, =4064x- inches 32768
Development of a Hand Tremor Quantification Device for the
Measurement of Pathalogical Tremor 46
36 y, = -8476 x - inches
32768 36
z,=-1148~- inches 32768
-
Chapter 4 Methodology
3. Displaving: motion in each sKis
For illustrative purposes the Tremor Quantifer displays the
motion of the tremor in each
of the three axes: x, y, and z. The pro- splits the data into
separate x, y, z arrays and
then displays the data in one of three formats:
1. raw data
2. mean position over acquisition period subtracted from
data
3. hi&-pass filtered data
1 . Raw Data
The K. y, z coordinates are graphed with respect to tirne. This
gives an indication of what
the raw movement looks like in each axes. An exarnple of raw
data c m be seen in
Figures 4.2 through 4.4 in which the raw data of an intention
tremor test is graphed in
each axis. Note that the y-coordinate of the graphs represents
distance (in cm) from the
miniBIRDTM transrnitter.
Figure 4.2: Movement of hand in x direction of intention
test
- - - - - - - - -- - - -
Development of a Hand Tremor Quantification Device for the
Measurement of Pathologicai Tremor 47