ECR2013_C-1123

Page 1 of 50

DTI tractography of lumbosacral plexus and sciatic nerve:normal FA and ADC values

Poster No.: C-1123

Congress: ECR 2013

Type: Scientific Exhibit

Authors: J. Broncano1, J. Etxano Cantera2, H. Samir1, J. M. Bondia2, J. L.

ZUBIETA2, J. D. Aquerreta2; 1Carmarthen/UK, 2Pamplona/ES

Keywords: Neuroradiology peripheral nerve, Musculoskeletal spine, MR-Diffusion/Perfusion, Imaging sequences

DOI: 10.1594/ecr2013/C-1123

Any information contained in this pdf file is automatically generated from digital materialsubmitted to EPOS by third parties in the form of scientific presentations. Referencesto any names, marks, products, or services of third parties or hypertext links to third-party sites or information are provided solely as a convenience to you and do not inany way constitute or imply ECR's endorsement, sponsorship or recommendation of thethird party, information, product or service. ECR is not responsible for the content ofthese pages and does not make any representations regarding the content or accuracyof material in this file.As per copyright regulations, any unauthorised use of the material or parts thereof aswell as commercial reproduction or multiple distribution by any traditional or electronicallybased reproduction/publication method ist strictly prohibited.You agree to defend, indemnify, and hold ECR harmless from and against any and allclaims, damages, costs, and expenses, including attorneys' fees, arising from or relatedto your use of these pages.Please note: Links to movies, ppt slideshows and any other multimedia files are notavailable in the pdf version of presentations.www.myESR.org

Page 2 of 50

Purpose

Sciatic nerve is the largest nerve of the body. It is originated by the ventral rami of L4to S3 spinal roots and exits the pelvis trough the greater sciatic foramen usually inferiorto the piriformis muscle, as distinct tibial and peroneal divisions, enclosed in a commonnerve sheath (1). It is one of the major neural pathways of the lower extremity allowingthe flexion of the knee by harmstring muscles and also provides all the sensory andmotor functions below the knee to the muscle groups innervated by its tibial and peronealdivisions (Fig. 1 on page 3) (1, 2).

Sciatica is one of the commonest variations of low back pain and is described as painradiating to the leg, normally below the knee and into the foot and toes (3-5). Themost common location of sciatic nerve and lumbosacral plexus pathology is the spine,due to lumbosacral hernia (1). The estimated prevalence, registered in the literature,rated sciatic symptoms between 1.2% and 43% (6). Peripheral entrapment can occurmore frequently in the pelvis. The main etiologies are trauma, iatrogenic injury -due togynecological or joint replacement surgeries- and the piriformis muscle syndrome (7).

MRI is a useful method to evaluate neural entrapment, especially by using T2 weightedsequences, like Short Time Inversion Recovery (STIR) T2 weighted images or fat-saturated heavily-T2 weighted sequences (8-12). However, in specific anatomicalregions such as lumbar nerve roots, conventional MR imaging has been inadequatefor evaluating symptomatic foraminal stenosis, because of the high incidence of falsepositives found in asymptomatic elderly patients (13, 14).

But with the appropriate magnetic field gradients, MRI could assess intrinsic tissularcharacteristic as well, like the random motion of water molecules inside anatomicstructures. This property, called diffusion, provides essential information about theultrastructure of the anatomic regions under study. When water motion is homogeneousin all directions in the space we talk about isotropic diffusion. In contrast, when thismovement is predominantly in one direction due to the intrinsic properties of the tissuebeing evaluated, like in nervous pathways, we talk about anisotropic diffusion (Fig. 2on page 4). Peripheral nerves have shown high degree of anisotropic diffusion dueto fibrillar alignment of axons with myelin sheaths and compartmentalization of the fiberbundles (15, 16).

Diffusion Tensor Imaging (DTI) is an MRI technique that using several weighteddiffusion planes acquisitions is able to establish the three-dimensional mayor directionsof water molecules movement (eigenvectors) and its magnitude (eigenvalues) owing tocalculate the grade of anisotropy by the Fractional Anisotropy (FA) and the Apparent

Page 3 of 50

Diffusion Coefficient (ADC). Different specifically design software traduces the FAregistered encoding the eigenvectors and its eigenvalues in a plane colour-coded maprepresentative of the structures under study. Fiber tracking represents these values in avirtual 3D colour-coded map (17, 18).

There are only two studies evaluating the sciatic nerve DTI- tractography in the thigh anda paucity of them assessing DTI parameters of lumbar nerve roots but, to the best ofour knowledge, there are no other works assessing the intrinsic properties of intrapelvicsciatic nerve and none of them establishing the normative values of sciatic nerve (14,16, 19-22).

The aim of our study was (1) to describe normal FA and ADC values of lumbosacralplexus and sciatic nerve and its relationship age, gender and anatomical variatons, (2)to evaluate the feasibility of using DTI-Tractography for the evaluation of lumbosacralplexus and sciatic nerve, in 1.5 T and 3 T MRI and (3) to demonstrate the reproducibilityof this imaging protocol among different radiologists.

Images for this section:

Page 4 of 50

Fig. 1

Fig. 2

Page 5 of 50

Methods and Materials

Patients

Thirty-two consecutive patients were prospectively enrolled, for realizing a pelvic MRIstudy done for other indication and adding the DTI-Tractography sequence to thestandard protocol without complications (Fig. 1 on page 8). The global mean agewas 45.19 ± 17.76 years. Sixteen patients were male (mean age of 47.44 ± 21.93 years)and the rest were female (mean age 42.94±12.65 years). Eighteen patients were scannedon 1,5 Tesla MRI (mean age 39.28 ± 16.74 years; 10 male and 8 female) and the otherfourteen on 3 Tesla MRI (mean age 52.79 ± 16.59 years, 6 male and 8 female).

As a result 32 patients and 64 sciatic nerves were registered with DTI and fiber tracking.The presence of any pathologic condition interesting the lumbosacral hinge or pelvicwaist, previous trauma or any neurological or systemic nosological entity that could because of sciatic neuropathy was considered as exclusion criteria. The institutional reviewboard approved the study protocol. All patients gave written informed consent before theexamination was done.

Acquisition Protocol

18 patients were evaluated in 1,5 Tesla MRI equipment (Symphony Tim, SiemensHealthcare, Erlangen, Germany) and the other 14 patients in 3 Tesla MRI (Trio Tim,Siemens Healthcare, Erlangen, Germany). All patients were studied in supine position.

The field gradient and slew rate of the 1.5 T MRI scan was 30mT/m2 and 125 T/m/s and

45 mT/m2 and 200 T/m/s for the 3 T MRI scan, respectively (Table 1 on page 11).

Page 6 of 50

Table 1References: Radiology, Glangwili General Hospital - Carmarthen/UK

Eight elements body phased array coil combined with high-resolution pelvic surface coilwere used. Coronal fat suppressed single-shot fast-spin-echo (FSE) echo-planar (EPI)DTI sequence with 30 encoded directions was used. In order to suppress background

body signal, long b diffusion weighted b value of 1000 s/mm2 and two b0 images per DTIsequence were applied. Image acquisition protocol is summarized in Table 1 on page11 .

The image acquisition parameters for the 1,5 T MRI scan were as follows: 9000/97 ms(TR/TE), 350 x 100 mm field of view (FoV), 2 mm slice thickness without gap, coronal

slice orientation with 2,9 x 2,9 x 2 mm3 calculated voxel size. Total acquisition sequencetime was 9 minutes 47 seconds. For 3 Tesla MRI scan, image acquisition parameterswere 8900/95 ms (TR/TE), 230 x 100 mm FoV, 2 mm slice thickness without gap and

a calculated voxel size of 1.9 x 1.9 x 2 mm3. The total sequence registration lasted 9minutes and 40 seconds. In both cases parallel imaging technique with a SENSE factorof 2, partial Fourier acquisition with a half-scan factor of 0,681, single-shot EPI factor

Page 7 of 50

of 122 and 1322 Hz per pixel frequency direction bandwidth were applied. Right-to-leftphase encoding and 122 x 122 matrix were also used (Table 1 on page 11).

Study evaluation

After DTI sequence acquisition, data were transferred to a workstation with and approvedvendor specific-designed software (Neuro3D, Siemens Healthcare, Erlangen, Germany)for quantitative assessment of DTI sequence and fiber tracking.

Colour-coded coronal maps were used to locate the spinal roots and sciatic nerve.Circular ROIs smaller than the nerves, for avoiding partial volume effects fromsurrounding tissues and vessels, were placed in the sciatic nerve and spinal roots at 5different levels: L4, L5 and S1 spinal roots, intrapelvic and extrapelvic portions of sciaticnerve (Fig. 3 on page 10). ROI location on the colour-coded coronal planes wereconfirmed using a T2 weighted Fast Spin Echo coronal sequence included in the standardMR protocol. Mean FA and ADC values were automatically calculated.

DTI tractography was performed by placing multiple seed points in a region larger thanthe nerve interested to visualize it, using an FA threshold of 0.1 and allowing fiberangulation of up to 30 degrees (Fig. 2 on page 9). Two radiologists reviewed all DTItractography studies and collected those data blindly.

Because FA and ADC values depend on signal-to-noise ratio (SNR) of diffusion tensoracquisition, we have determined the SNR for all DTI sequences, using the followingequation:

Fig. 4References: Radiology, Glangwili General Hospital - Carmarthen/UK

SInerve represents the mean signal intensity within a ROI (mean size of 0,1 cm2) placed onthe intrapelvic portion of the sciatic nerve. SDbackground_noise denote the average standard

deviation of the background noise measured by four standardised ROIs rating 3 cm2

placed in four reproducible image locations outside the pelvis. A correction factor wasused to account for the systematic error in noise measurements in magnitude images.These SNR values were determined on 1,5 T and 3 T studies to verify the possible

Page 8 of 50

extrapolation of the quantitative diffusion and anisotropy parameters between the twotypes of MRI scans (23-25).

Statistical analysis

A two tailed alpha evidence level of 0.05 was set previous to the realization of statisticalstudies using a commercially available statistical program (SPSS version 17.0, SPSSInc.). Also, exact p value calculation was applied. The kolmogorov-Smirnov test wasused to corroborate the normality distribution of the study sample. Due to the smallsample size and absence of normal distribution of some variables, comparison of theSNR, FA and ADC values between patient genre and type of MRI scan; U Mann-Withneynon-parametric tests were applied. In the evaluation of the intrasubject side-to-side andanatomic location variability of the FA, ADC and SNR values T of Wilcoxon test weredone.

Patients were divided, according to their age, into three groups: 18 - 34 years, 35 -49 years and those with at least 50 years. To compare DTI quantitative measures andSNR between different age groups, Kruskal-Wallis test was used. Intraclass correlationcoefficient (ICC) and Bland-Altman curves were assessed in order to evaluate theinterobserver variability and, also, to graphically represent intrasubject evaluation of SNR.ICCs were interpreted according to the criteria of Landis and Koch (26): ICC of 0.01 -0,20 indicated slight agreement, ICCs of 0.21-0.40, fair agreement, ICCs of 0.41-0.60,moderate agreement, ICCs of 0.61 - 0.80, substantial agreement and ICC of 0.81 - 1.0,almost excellent agreement.

All the quantitative measurements were represented using median and interquartileranges. 95% confident intervals were calculated in all the bar diagrams included in thisstudy.

Images for this section:

Page 9 of 50

Fig. 1

Page 10 of 50

Fig. 2

Page 11 of 50

Fig. 3

Fig. 4

Page 12 of 50

Table 1

Page 13 of 50

Results

Patient sex and side-to-side variability

The mean FA and ADC for L4, L5 and S1 nerve roots, intrapelvic and extrapelvic portionsof the sciatic nerve were 0.45 ± 0.11, 0.44 ± 0.11, 0.41 ± 0.11, 0.49 ± 0.06 and 0.49 ±

0.12, for FA and 1.213 ± 0.295 mm2/s, 1.275 ± 0.243 mm2/s, 1.357 ± 0.236 mm2/s, 1.252

± 0.342 mm2/s and 1.251 ± 0.246 mm2/s for ADC, respectively (Table 2 on page 18).

Table 2References: Radiology, Glangwili General Hospital - Carmarthen/UK

For the evaluation of the intrasubject side-to-side variability of the sciatic nerve diffusionparameters, T of Wilcoxon tests were done to compare the FA and ADC values betweenthe 64 sciatic nerves under evaluation. The results of these comparisons are representedon Table 2 on page 18 and no significant variations were observed (p>0.05).

Regarding the quantitative DTI evaluation of L4, L5 and S1 nerve roots and sciatic nervevariations between male and female patients, no substantial variations among them wereregistered. (p>0.05, see Table 3 on page 19, Table 4 on page 19 and Fig. 5 onpage 21).

Page 14 of 50

Table 3References: Radiology, Glangwili General Hospital - Carmarthen/UK

Table 4References: Radiology, Glangwili General Hospital - Carmarthen/UK

Anatomical variances

A progressive decrease on FA and ADC values from cranial to lower nerve roots and anincrease of those parameters from proximal to distal portions of the sciatic nerve werestated on descriptive values being, those differences statistically significant between L4

and S1 nerve roots (0.45 ± 0.11 vs. 0.41 ± 0.11, p=0.013; 1.213 ± 0.295 mm2/s vs. 1.357

± 0.236 mm2/s, p<0.001, for FA and ADC of L4 and S1 nerve roots, respectively). The

Page 15 of 50

same results were obtained in the comparison of L5 spinal root and intrapelvic portion of

the sciatic nerves (0.44 ± 0.11 vs. 0.49 ± 0.06, p<0.001; 1.275 ± 0.243 mm2/s vs. 1.252

± 0.342 mm2/s, p=0.027 for FA and ADC of L5 nerve root and intrapelvic portion of thesciatic nerve; Table 2 on page 18 and Fig. 6 on page 22).

Fig. 6References: Radiology, Glangwili General Hospital - Carmarthen/UK

Contrarily, no significant differences were evoked by statistical analyses in thecomparison of FA and ADC between intrapelvic and extrapelvic portions of sciatic nerve

(0.49 ± 0.06 vs. 0.49 ± 0.12, p=0.878; 1.252 ± 0.342 mm2/s vs. 1.251 ± 0.246 mm2/s,p=0.647, respectively; Table 2 on page 18 and Fig. 6 on page 22 ).

DTI tractography and MRI type of scan

Centering on the evaluation of quantitative DTI measures between 1.5 T and 3 T MRscans, the FA and ADC values of L4, L5 and S1 nerve roots, intrapelvic and extrapelvic

Page 16 of 50

portions of sciatic nerve did not show significant differences between 1.5 and 3 TeslaMRI scanners (p>0.05; see Table 5 on page 20; Fig. 7 on page 23).

Table 5References: Radiology, Glangwili General Hospital - Carmarthen/UK

DTI tractography and patients age

According to patient's age and between the three groups generated for this analysis(18-34 years, n=10 patients; 34-49 years, n=10 patients and more than 50 years, n=12patients), Kruskal-Wallis tests did not show significant differences in the FA and ADCvalues of L4, right L5 and S1 nerve roots, intrapelvic and extrapelvic portion of the sciaticnerve (p>0.05). Substantial variations among three age groups were observed in left L5nerve root (0.40 ± 0.11, 0.39 ± 0.04 and 0.50 ± 0.06 for 18-34, 35-59 and >50 yearsrespectively, p=0.006; Table 6 on page 20).

Page 17 of 50

Table 6References: Radiology, Glangwili General Hospital - Carmarthen/UK

DTI signal-to-noise ratio, patient gender and age, intrasubject variability and MRIscan

No significant differences (p>0.05) were obtained in the mean left, right and global signal-to-noise ratio (SNR) between male and female patients and, also, regarding the type ofMRI scans under consideration (Table 7 on page 21 and Fig. 8 on page 24).

Page 18 of 50

Table 7References: Radiology, Glangwili General Hospital - Carmarthen/UK

Absence of intrasubject variability were registered among SNR (23.23 ± 13.72 vs. 20.60± 16.94 for right and left SNR respectively; p=0.674), as we could see in the Bland Altmanplot presented in Table 7 on page 21 and Fig. 9 on page 25. Finally, no substantialvariations were observed in the SNR between different age groups (p>0.05; Table 7 onpage 21).

Interobserver agreement

For assessing the reproducibility of the DTI evaluation made by the two blind observersof the lumbosacral plexus and sciatic nerve FA and ADC values, intraclass correlationcoefficient (ICC) were calculated. An excellent correlation, according to Landis and Kochcriteria, were obtained in the quantitative evaluation of ADC and FA (ICC=0.803) of L4,L5 and S1 nerve roots and sciatic nerve between the two readers.

Images for this section:

Page 19 of 50

Table 2

Table 3

Page 20 of 50

Table 4

Table 5

Page 21 of 50

Table 6

Table 7

Page 22 of 50

Fig. 5

Page 23 of 50

Fig. 6

Page 24 of 50

Fig. 7

Page 25 of 50

Fig. 8

Page 26 of 50

Fig. 9

Page 27 of 50

Conclusion

Originated by the union the ventral rami of L4 to S3 spinal roots, sciatic nerve is thelargest nerve of the body (27). Exits the pelvis trough the greater sciatic foramen usuallyinferior to the piriformis muscle, as distinct tibial and peroneal divisions, enclosed in acommon nerve sheath (1, 27). But, in some circumstances and, specially, the peronealdivision could exit the pelvis across the muscle. The sciatic nerve continues descendingthrough the thigh, posterior to adductor magnus muscle and anterior to glutaeus maximusmuscle. In the distal segment of the thigh, the two divisions are physically separated intibial and common peroneal nerves (Fig. 1 on page 33) (7, 27).

In the pelvis, sciatic nerve innervates piriformis and quadratus femoris muscles. In thethigh, the tibial nerve innervates the caput longum of the biceps femoris muscle and also,semitendinosus, semimembranosus and adductor magnus muscles. Peroneal divisioninnervates the caput brevis of the biceps femoris muscle (28). It is one of the major neuralpathways of the lower extremity allowing the flexion of the knee by harmstring musclesand also provides all the sensory and motor functions below the knee (1, 2).

Diffusion tensor imaging (DTI) can obtain valuable information regarding themicrostructure of tissues by monitoring the random movement of water molecules, whichis usually restricted in anisotropic tissues (23, 29-35). DTI is specially advantageous fortissues containing organized microstructure, such as white matter tracts in the brain.In such bundles, water molecules move predominantly along the fibers than in otherdirections, and this is called anisotropic diffusion. If there is no directional variation ratein tissues, diffusion is said to be isotropic (Fig. 2 on page 34; 14, 32). Consequently,by applying different motion probe gradients in some directions, DTI is able to establishthe three-dimensional mayor directions of water molecules movement (eigenvectors) andits magnitude (eigenvalues). The direction of the largest eigenvalue correlates with thedirection of the largest diffusion (14, 16).

Apparent diffusion coefficient (ADC) is a scalar value measured in mm2/s,independent of magnetic field strength and reflects molecular diffusivity under motionrestriction (23, 32, 36, 37). Fractional anisotropy (FA) is a quantitative index used tocharacterize directional variability in diffusion (23). Both FA and ADC values depend onthe acquisition technique and other factors such as the number of gradient directions,voxel size and number of acquisitions (19). FA values range from 0 to 1, with FA valuesproximal to 1 indicating anisotropic diffusion and values close to 0 seen in tissues withisotropic diffusion.

Page 28 of 50

Different specifically design software traduces the FA registered encoding theeigenvectors and its eigenvalues in a plane colour-coded map representative of thestructures under study (17-19). Fiber tracking is an algorithm constructing a fiber fromthe tensors using both shape (anisotropy) and orientation characteristics (Fig. 2 on page34). A starting point is selected and a propagation line (fiber) is generated if thepredefined criteria for iterative propagation, such as largest fiber propagation, minimumfiber length, step size and anisotropy threshold, are fullfiled (21). The number of tractsvisualized by DTI did not present the actual volume of fiber trajectories (14).

Since Skorpil et al. published the first DTI peripheral nerve tractography of the sciaticnerve at the thigh, this study protocol has been successfully applied in the assessment ofother proximal and distal peripheral nerves in human beings, like median, ulnar and radialnerves in the upper limb and the thigh course of sciatic nerve, tibial and peroneal nervesin the lower extremity and, also, in non-compressed and compressed lumbar nerve roots(5, 14-16, 19-23, 32, 38-41).

However, this is the first time, to best of our knowledge, that the lumbosacral plexusand sciatic nerve, in its intrapelvic and extrapelvic portions, have been scanned duringthe same acquisition. In those studies the field of view (FoV) orientation was the axialplane, right perpendicular to the major axle of the limb being evaluated. In contrast, theplanning of the FoV in our study was using a coronal plane with right to left phase encodeddirection acquisition. This technical strategy permits, without a substantial increasing ofthe number of slices and, therefore, acquisition time, to acquire the entire sacral plexus,the intrapelvic and extrapelvic portions of the sciatic nerve nearly at the same plane withno variation on the colour codification system (Table 1 on page 35).

Page 29 of 50

Table 1References: Radiology, Glangwili General Hospital - Carmarthen/UK

The FA and ADC values of the lumbar nerve roots obtained in our study (Table 2 on page36) are slightly higher than other previously published of the lumbar nerve roots andmedular cord (14, 19, 20). But, on the other hand, are weakly lower than those obtainedin other parts of the anatomy such as the median nerves (0.45-0.6) (15, 23, 39). Thesedifferences could be related to the b value and number of encoded-directions applied,coil configuration, the phase-encoded direction set in the sequence and the anatomicaluniqueness of the sciatic nerve, regarding its dimension and also, the presence of twomajor divisions enclosed in a common nerve sheath.

On fiber tracking, lumbar nerve roots were coded in blue (cephalocaudal direction) inthe lateral recess and purple in the foramen (cephalocaudal direction and mediolateralorientation), as described by Balbi et al (19). Moreover, the intrapelvic portion of thesciatic nerve was coded in purple too, because the oblique right-left and crania-caudalorientation and, after exiting the pelvis through the greater sciatic foramen, it turned intoa blue colour-coded neural pathway (Fig. 1 on page 33, Fig. 2 on page 34).

Page 30 of 50

As published by other authors in different anatomical locations, no side-to side variabilityand male-to-female differences were obtained during the analysis of ADC and FA valuesof L4, L5, S1 nerve roots and sciatic nerve (15, 19, 39). This finding reinforces the useof the contralateral side as an internal control in all patients for whom DTI is used in theevaluation of unilateral peripheral neuropathies interesting the sciatic nerve (Table 2 onpage 36,Table 3 on page 36 andTable 4 on page 37; Fig. 5 on page 37and Fig. 7 on page 39).

Table 3References: Radiology, Glangwili General Hospital - Carmarthen/UK

Table 4References: Radiology, Glangwili General Hospital - Carmarthen/UK

Page 31 of 50

Regarding quantitative DTI anatomical variation inside the same neural pathway,Guggenberger et al. found that normative diffusion values and cross sectional area of themedian nerve increased from proximal to distal in healthy individuals, and was greater inpatients affected with carpal tunnel syndrome than in healthy individuals (15). Significantanatomical decrease in FA and ADC values were observed in our study comparing L4and S1 spinal nerve roots and an increase of them between L5 nerve root and intrapelvicportion of the sciatic nerve. But, contrarily, no substantial variations were registeredcomparing intrapelvic and extrapelvic portions of the sciatic nerves (Table 2 on page36,Table 3 on page 36 and Table 4 on page 37; Fig. 6 on page 38).

Fig. 6References: Radiology, Glangwili General Hospital - Carmarthen/UK

These differences may be sustained to the nerve section and other anatomicalaspects as there were no significant differences between the two portions of thesciatic nerves, regions with identical anatomical configuration regarding the differencein fiber orientation. Moreover, the significant difference between L5 nerve root andintrapelvic portion of sciatic nerve may be explained by histological differences: less

Page 32 of 50

collagen contained in spinal nerve root as in peripheral nerve endoneurium, absenceof perineurium and less developed epineurium in spinal nerve roots than in peripheralnerves (19, 42, 43).

Considering the influence of age with FA and ADC values of median nerve DTI,Guggenberger et al. registered similar behavior of those parameters to age similarly tothe white matter of the brain. In this situation, diffusion changes of peripheral nerves mightprimarily reflect changes in the degree of myelinization, as shown by autopsy studies inwich degeneration of myelin sheats, axon deletion, rupture of intracellular microtubuli andaugmentation of the interstitium with increasing age were described. And these findingscould lead to increased diffusitivity and decreased FA values (15, 44-48). In our studyno significant variation, except for FA of left L5 nerve root, were obtained (Table 6 onpage 40). This result could be due to the presence of some outliers and, also, de lownumber of subjects contained in each age group. Further investigation must be done toestablish properly the relationship between age related changes of the lumbar spine andquantitative DTI evaluation of lumbosacral plexus and sciatic nerve.

Considering the type of MRI scan used no significant differences were seen between 1.5Tesla and 3 Tesla magnetic fields in FA, ADC and SNR values (Table 5 on page 41and Table 7 on page 41, Fig. 7 on page 39). Also, no substantial variations inSNR were registered between male and female patients and right-to-left side (Table 7on page 41, Fig. 8 on page 42 and Fig. 9 on page 43). These results revealthat, although the imaging of spinal cord and peripheral nerve is challenging becauseof technical limitations such as the inherent low SNR of EPI sequences, small size oflumbar nerve roots, susceptibility artifacts because of tissue-bone interfaces and themotion artifacts arising from respiratory activity, DTI tractography of the sciatic nerve isfeasible with good image quality.

As reported for median nerve DTI (15, 41), FA and ADC values agreement between thetwo blinded radiologist readers were excellent (ICC=0.804) indicating that quantitativeevaluation of diffusion properties of the sciatic nerve is reproducible between two differentobservers. This interreader agreement has not been reported in other studies concerninglumbar nerve roots evaluation (14, 19, 20).

We acknowledge that our study has several limitations. The first is that a small numberof subjects were investigated in order to accurately ascertan the relationship between FAand ADC values of the sciatic nerve and patient's age. Secondly, we did not compare theADC and FA values of lumbar nerve roots and sciatic nerves between healthy individualsand patients with spinal or extra spinal sciatic nerve compression for establishing anADC and FA thresholds to diagnose sciatic neuropathy. Lastly, although assessinginterreader variability, absence of intrareader comparison for assessing the precision ofDTI tractography of lumbosacral plexus and sciatic nerve was done. For that reason

Page 33 of 50

further investigation is needed in order to achieve a profound knowledge regardingfunctional imaging of one of the largest nerves of the body.

In conclusion, DTI tractography is a feasible and reproducible technique for assessingthe anatomy, orientation and integrity of non-pathological lumbosacral plexus andsciatic nerve, by means of FA and ADC. Those normative values could be usedas a reference for the assessment of suspected sciatic nerve injuries. Abscence ofage related changes and presence of anatomical variations of the sciatic nerve areconfirmed, and may be taked into account during the evaluation of it.

Fig. 11References: Radiology, Glangwili General Hospital - Carmarthen/UK

Images for this section:

Page 34 of 50

Fig. 1

Page 35 of 50

Fig. 2

Page 36 of 50

Table 1

Table 2

Page 37 of 50

Table 3

Table 4

Page 38 of 50

Fig. 5

Page 39 of 50

Fig. 6

Page 40 of 50

Fig. 7

Page 41 of 50

Table 6

Table 7

Page 42 of 50

Table 5

Page 43 of 50

Fig. 8

Fig. 9

Page 44 of 50

References

1. Petchprapa CN, Rosenberg ZS, Sconfienza LM, Cavalcanti CF, Vieira RL, Zember JS.MR imaging of entrapment neuropathies of the lower extremity. Part 1. The pelvis andhip. Radiographics : a review publication of the Radiological Society of North America,Inc. 2010;30(4):983-1000.

2. Lanzieri CF, Hilal SK. Computed tomography of the sacral plexus and sciatic nerve inthe greater sciatic foramen. AJR American journal of roentgenology. 1984;143(1):165-8.

3. Fairbank JC. Sciatic: An archaic term. BMJ. 2007;335(7611):112.

4. van Tulder M, Koes B, Bombardier C. Low back pain. Best practice & research Clinicalrheumatology. 2002;16(5):761-75.

5. Konstantinou K, Dunn KM. Sciatica: review of epidemiological studies and prevalenceestimates. Spine. 2008;33(22):2464-72.

6. Videman T, Nurminen T, Tola S, Kuorinka I, Vanharanta H, Troup JD. Low-back painin nurses and some loading factors of work. Spine. 1984;9(4):400-4.

7. Benzon HT, Katz JA, Benzon HA, Iqbal MS. Piriformis syndrome: anatomicconsiderations, a new injection technique, and a review of the literature. Anesthesiology.2003;98(6):1442-8.

8. Bendszus M, Stoll G. Technology insight: visualizing peripheral nerve injury using MRI.Nature clinical practice Neurology. 2005;1(1):45-53.

9. Chappell KE, Robson MD, Stonebridge-Foster A, Glover A, Allsop JM, Williams AD,et al. Magic angle effects in MR neurography. AJNR American journal of neuroradiology.2004;25(3):431-40.

10. Howe FA, Saunders DE, Filler AG, McLean MA, Heron C, Brown MM, et al.Magnetic resonance neurography of the median nerve. The British journal of radiology.1994;67(804):1169-72.

11. Stoll G, Bendszus M, Perez J, Pham M. Magnetic resonance imaging of the peripheralnervous system. Journal of neurology. 2009;256(7):1043-51.

12. Zhang Z, Song L, Meng Q, Li Z, Pan B, Yang Z, et al. Morphological analysisin patients with sciatica: a magnetic resonance imaging study using three-dimensionalhigh-resolution diffusion-weighted magnetic resonance neurography techniques. Spine.2009;34(7):E245-50.

Page 45 of 50

13. Aota Y, Niwa T, Yoshikawa K, Fujiwara A, Asada T, Saito T. Magnetic resonanceimaging and magnetic resonance myelography in the presurgical diagnosis of lumbarforaminal stenosis. Spine. 2007;32(8):896-903.

14. Eguchi Y, Ohtori S, Orita S, Kamoda H, Arai G, Ishikawa T, et al. Quantitativeevaluation and visualization of lumbar foraminal nerve root entrapment by usingdiffusion tensor imaging: preliminary results. AJNR American journal of neuroradiology.2011;32(10):1824-9.

15. Guggenberger R, Nanz D, Puippe G, Rufibach K, White LM, Sussman MS, etal. Diffusion tensor imaging of the median nerve: intra-, inter-reader agreement, andagreement between two software packages. Skeletal radiology. 2012;41(8):971-80.

16. Skorpil M, Engstrom M, Nordell A. Diffusion-direction-dependent imaging: anovel MRI approach for peripheral nerve imaging. Magnetic resonance imaging.2007;25(3):406-11.

17. Mukherjee P, Berman JI, Chung SW, Hess CP, Henry RG. Diffusion tensor MRimaging and fiber tractography: theoretic underpinnings. AJNR American journal ofneuroradiology. 2008;29(4):632-41.

18. Mukherjee P, Chung SW, Berman JI, Hess CP, Henry RG. Diffusion tensor MRimaging and fiber tractography: technical considerations. AJNR American journal ofneuroradiology. 2008;29(5):843-52.

19. Balbi V, Budzik JF, Duhamel A, Bera-Louville A, Le Thuc V, Cotten A. Tractographyof lumbar nerve roots: initial results. European radiology. 2011;21(6):1153-9.

20. Eguchi Y, Ohtori S, Yamashita M, Yamauchi K, Suzuki M, Orita S, et al. Clinicalapplications of diffusion magnetic resonance imaging of the lumbar foraminal nerve rootentrapment. European spine journal : official publication of the European Spine Society,the European Spinal Deformity Society, and the European Section of the Cervical SpineResearch Society. 2010;19(11):1874-82.

21. Skorpil M, Karlsson M, Nordell A. Peripheral nerve diffusion tensor imaging. Magneticresonance imaging. 2004;22(5):743-5.

22. Takahara T, Hendrikse J, Kwee TC, Yamashita T, Van Cauteren M, Polders D, etal. Diffusion-weighted MR neurography of the sacral plexus with unidirectional motionprobing gradients. European radiology. 2010;20(5):1221-6.

23. Andreisek G, White LM, Kassner A, Sussman MS. Evaluation of diffusion tensorimaging and fiber tractography of the median nerve: preliminary results on intrasubjectvariability and precision of measurements. AJR American journal of roentgenology.2010;194(1):W65-72.

Page 46 of 50

24. Dietrich O, Heiland S, Sartor K. Noise correction for the exact determination ofapparent diffusion coefficients at low SNR. Magnetic resonance in medicine : officialjournal of the Society of Magnetic Resonance in Medicine / Society of MagneticResonance in Medicine. 2001;45(3):448-53.

25. Henkelman RM. Measurement of signal intensities in the presence of noise in MRimages. Medical physics. 1985;12(2):232-3.

26. Landis JR, Koch GG. The measurement of observer agreement for categorical data.Biometrics. 1977;33(1):159-74.

27. Cejas C, Aguilar M, Falcon L, Caneo N, Acuna MC. High resolution (3T) magneticresonance neurography of the sciatic nerve. Radiologia. 2012.

28. Byun WM, Jang HW, Kim SW. Three-dimensional magnetic resonance renderingimaging of lumbosacral radiculography in the diagnosis of symptomatic extraforaminaldisc herniation with or without foraminal extension. Spine. 2012;37(10):840-4.

29. Bammer R. Basic principles of diffusion-weighted imaging. European journal ofradiology. 2003;45(3):169-84.

30. Bammer R, Acar B, Moseley ME. In vivo MR tractography using diffusion imaging.European journal of radiology. 2003;45(3):223-34.

31. Basser PJ, Mattiello J, LeBihan D. MR diffusion tensor spectroscopy and imaging.Biophysical journal. 1994;66(1):259-67.

32. Hiltunen J, Suortti T, Arvela S, Seppa M, Joensuu R, Hari R. Diffusiontensor imaging and tractography of distal peripheral nerves at 3 T. Clinicalneurophysiology : official journal of the International Federation of ClinicalNeurophysiology. 2005;116(10):2315-23.

33. Mori S, Barker PB. Diffusion magnetic resonance imaging: its principle andapplications. The Anatomical record. 1999;257(3):102-9.

34. Mori S, van Zijl PC. Fiber tracking: principles and strategies - a technical review. NMRin biomedicine. 2002;15(7-8):468-80.

35. Nucifora PG, Verma R, Lee SK, Melhem ER. Diffusion-tensor MR imagingand tractography: exploring brain microstructure and connectivity. Radiology.2007;245(2):367-84.

36. Hein PA, Eskey CJ, Dunn JF, Hug EB. Diffusion-weighted imaging in the follow-upof treated high-grade gliomas: tumor recurrence versus radiation injury. AJNR Americanjournal of neuroradiology. 2004;25(2):201-9.

Page 47 of 50

37. Weber J, Mattle HP, Heid O, Remonda L, Schroth G. Diffusion-weighted imaging inischaemic stroke: a follow-up study. Neuroradiology. 2000;42(3):184-91.

38. Andreisek G, White LM, Kassner A, Tomlinson G, Sussman MS. Diffusion tensorimaging and fiber tractography of the median nerve at 1.5T: optimization of b value.Skeletal radiology. 2009;38(1):51-9.

39. Kabakci N, Gurses B, Firat Z, Bayram A, Ulug AM, Kovanlikaya A, et al. Diffusiontensor imaging and tractography of median nerve: normative diffusion values. AJRAmerican journal of roentgenology. 2007;189(4):923-7.

40. Khalil C, Budzik JF, Kermarrec E, Balbi V, Le Thuc V, Cotten A. Tractography ofperipheral nerves and skeletal muscles. European journal of radiology. 2010;76(3):391-7.

41. Khalil C, Hancart C, Le Thuc V, Chantelot C, Chechin D, Cotten A. Diffusion tensorimaging and tractography of the median nerve in carpal tunnel syndrome: preliminaryresults. European radiology. 2008;18(10):2283-91.

42. Ellingson BM, Ulmer JL, Kurpad SN, Schmit BD. Diffusion tensor MR imaging ofthe neurologically intact human spinal cord. AJNR American journal of neuroradiology.2008;29(7):1279-84.

43. Olmarker K, Rydevik B. Pathophysiology of sciatica. The Orthopedic clinics of NorthAmerica. 1991;22(2):223-34.

44. Meier-Ruge W, Ulrich J, Bruhlmann M, Meier E. Age-related white matter atrophy inthe human brain. Annals of the New York Academy of Sciences. 1992;673:260-9.

45. Miller JH, McKinstry RC, Philip JV, Mukherjee P, Neil JJ. Diffusion-tensor MR imagingof normal brain maturation: a guide to structural development and myelination. AJRAmerican journal of roentgenology. 2003;180(3):851-9.

46. Peters A. The effects of normal aging on myelin and nerve fibers: a review. Journalof neurocytology. 2002;31(8-9):581-93.

47. Pfefferbaum A, Sullivan EV. Increased brain white matter diffusivity in normal adultaging: relationship to anisotropy and partial voluming. Magnetic resonance in medicine :official journal of the Society of Magnetic Resonance in Medicine / Society of MagneticResonance in Medicine. 2003;49(5):953-61.

48. Verdu E, Ceballos D, Vilches JJ, Navarro X. Influence of aging on peripheralnerve function and regeneration. Journal of the peripheral nervous system : JPNS.2000;5(4):191-208.

Page 48 of 50

Personal Information

Fig. 11References: Radiology, Glangwili General Hospital - Carmarthen/UK

Images for this section:

Page 49 of 50

Fig. 11

Page 50 of 50