Carpal Tunnel Syndrome: Diagnosis by Means of Median Nerve Elasticity—Improved Diagnostic Accuracy...

Original research n

Musculoskeletal IM

agIng

Radiology: Volume 270: Number 2—February 2014 n radiology.rsna.org 481

carpal Tunnel syndrome: Diagnosis by Means of Median Nerve Elasticity—Improved Diagnostic Accuracy of US with Sonoelastography1

Hideaki Miyamoto, MDEthan J. Halpern, MDMartin Kastlunger, MDMarkus Gabl, MDRohit Arora, MDRosa Bellmann-Weiler, MDGudrun M. Feuchtner, MD, PhDWerner R. Jaschke, MD, PhDAndrea S. Klauser, MD

Purpose: To compare the elasticity of the median nerve (MN) be-tween healthy volunteers and patients with carpal tunnel syndrome (CTS) and to evaluate the diagnostic utility of sonoelastographic measurements of the elasticity of the MN.

Materials and Methods:

This study was performed with institutional review board approval and written informed consent from all partici-pants. Hands in 22 healthy volunteers and in 31 patients with symptomatic CTS were studied. The cross-sectional area (CSA) and the elasticity of the MN, which was mea-sured as the acoustic coupler (AC)/MN strain ratio, were evaluated.

Results: Both hands in 22 healthy volunteers (three men [mean age, 52.7 years; age range, 41–65 years]; 19 women [mean age, 62.2 years; age range, 40–88 years]) and 43 hands in 31 patients with symptomatic CTS (three men [mean age, 69.0 years; age range, 46–88 years]; 28 women [mean age, 61.2 years; age range, 39–92 years]) were studied. Both the AC/MN strain ratio and the CSA in the patients with CTS were significantly higher than those in the healthy volunteers (P , .001). The presence of CTS was predicted by means of AC/MN strain ratio and CSA cutoff values, respectively, of 4.3 and 11 mm2, with areas under the receiver operating characteristic curves (AUCs) of 0.78 (95% confidence interval [CI]: 0.69, 0.88) and 0.85 (95% CI: 0.78, 0.93). A logistic model that combined the AC/MN strain ratio and the CSA improved diagnostic accuracy for CTS, with an AUC of 0.91 (95% CI: 0.85, 0.97; P , .001).

Conclusion: Sonoelastography provides significant improvement in the diagnostic accuracy of the ultrasonographic assessment of CTS.

q RSNA, 2013

1 From the Department of Diagnostic Radiology (H.M., M.K., G.M.F., W.R.J., A.S.K.), Department of Trauma Surgery and Sports Medicine (M.G., R.A.), and Department of Internal Medicine I, Clinical Immunology and Infectious Diseases (R.B.), Medical University Innsbruck, Anichstrasse 35, A-6020 Innsbruck, Austria; Department of Orthopaedic Surgery, Graduate School of Medicine, University of Tokyo, Tokyo, Japan (H.M.); and Department of Radiology, Thomas Jefferson University, Philadelphia, Pa (E.J.H.). Received December 31, 2012; revision requested February 11, 2013; revision received July 12; accepted July 23; final version accepted August 7. Address correspondence to A.S.K. (e-mail: [email protected]).

q RSNA, 2013

Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.

482 radiology.rsna.org n Radiology: Volume 270: Number 2—February 2014

MUSCULOSKELETAL IMAGING: Carpal Tunnel Syndrome: Diagnosis by Means of Median Nerve Elasticity Miyamoto et al

region of interest as a translucent color-coded, real-time image superimposed on the B-mode image. The color code indi-cated the relative stiffness of the tissues within the region of interest and ranged from red (soft) to blue (hard). Green and yellow indicated medium elasticity. The strain indicator on the lateral part in the screen indicated whether the displace-ment was sufficient to obtain local strains within the region of interest. The elasto-grams were constructed automatically by using the same optimal settings through-out the study, as previously suggested by Havre et al (14). Images were stored as cine loops in the memory of the US system. Representative sonoelastograph-ic images were chosen from the middle of the compression-relaxation cycle. Our sonoelastographic equipment can calcu-late the resultant strain ratio between any two areas as an index of elasticity. The elasticity of the MN was assessed as the AC/MN strain ratio. The measurements were repeated three times, and the resul-tant average strain ratio was calculated.

As a preliminary examination, we investigated interobserver agreement for the measurement of the AC/MN strain ratio. Exclusion criteria for this preliminary examination included the presence of a scar in the wrist from

distribution elastogram, estimated Young modulus, or semiquantitative values for strain ratios. With use of the semiquan-titative values for strain ratios, sonoelas-tography has demonstrated promising preliminary results for the diagnosis of masses of the liver, breast, pancreas, prostate, and thyroid with defined cut-off values (9–13). The aims of this study were to compare the elasticity of the MN between healthy volunteers and patients with CTS and to evaluate the diagnostic usefulness of the elasticity of the MN measured with sonoelastography.

Materials and Methods

This study was performed with the ap-proval of the institutional review boards of our institution (Medical University Innsbruck, Innsbruck, Austria). All participants provided written informed consent.

US was performed by using a 5- to 18-MHz linear array transducer (HI VI-SION Preirus; Hitachi-Aloka Medical, Tokyo, Japan). We attached an acoustic coupler (AC) (EZU-TECPL1; Hitachi-Alo-ka Medical) with a standardized elasticity to the transducer as a reference medium. Each subject was asked to sit in front of an examination table with his or her elbow flexed at 90° and the hand supi-nated. Fingers were kept relaxed, and a slight flexion of the wrist was maintained during the measurement. The proximal inlet of the carpal tunnel at the scaphoid-pisiform level was scanned in a trans-verse plane. Using conventional B-mode US, we measured the MN CSA by tracing a continuous line within the hyperechoic boundary of the nerve.

Sonoelastographic images were obtained by means of repeated com-pressions of the palm with the probe. Compression-decompression cycles were performed, with force and frequency ad-justed to an appropriate range according to the strain indicator on the screen. The elastogram appeared within a rectangular

Carpal tunnel syndrome (CTS) is a common compression neuropathy of the median nerve (MN) at the

level of the wrist. The prevalence of CTS has been estimated to be 50 cases per 1000 subjects per year (1). The main symptoms of CTS include numb-ness and tingling in the area of the MN distribution and weakness of the oppos-ing thumb (1).

There is no standard of reference for establishing a diagnosis of CTS. However, the general approaches for diagnosing the condition are the clinical provocation test and the nerve conduction velocity (NCV) test (1). B-mode ultrasonography (US) has been proposed as an aid in the initial assessment of CTS (2). The most com-monly agreed on B-mode US finding for the diagnosis of CTS is the enlargement of the MN cross-sectional area (CSA), for which cutoff values range from 8.5 to 12 mm2 (3–7). In a meta-analysis, the single test accuracy of CSA for CTS had 87% sensitivity and 83% specificity (8).

Sonoelastography is a modality for assessing the elasticity of soft tissue that uses conventional B-mode US with a color

Implication for Patient Care

n Sonoelastography can signifi-cantly improve the US assess-ment of CTS.

Advances in Knowledge

n The elasticity of the median nerve in patients with carpal tunnel syndrome (CTS), which was measured as the acoustic coupler–median nerve strain ratio, was significantly stiffer than that in healthy volunteers (P , .001).

n The presence of CTS was pre-dicted by means of a strain ratio cutoff value of 4.3, resulting in a sensitivity of 82% (95% confi-dence interval [CI]: 66%, 91%), at a specificity of 68% (95% CI: 53%, 80%), with an area under the receiver operating character-istic curve of 0.78 (95% CI: 0.69, 0.88).

n A logistic model combining strain ratio with the cross-sectional area (CSA) of the median nerve resulted in improved diagnostic accuracy for CTS compared with use of either CSA or the strain ratio alone.

Published online before print10.1148/radiol.13122901 Content codes:

Radiology 2014; 270:481–486

Abbreviations:AC = acoustic couplerCI = confidence intervalCSA = cross-sectional areaCTS = carpal tunnel syndromeICC = intraclass correlation coefficientMN = median nerveNCV = nerve conduction velocityROC = receiver operating characteristic

Author contributions:Guarantors of integrity of entire study, H.M., A.S.K.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manu-script revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, H.M., M.K., R.B., G.M.F., A.S.K.; clinical studies, H.M., M.K., M.G., R.A., R.B., G.M.F., A.S.K.; statistical analysis, H.M., E.J.H., M.K., R.B., G.M.F., A.S.K.; and manu-script editing, H.M., E.J.H., M.K., M.G., R.B., W.R.J., A.S.K.

Conflicts of interest are listed at the end of this article.



ICC was classified as poor (0.00–0.20), fair (0.21–0.40), good (0.41–0.75), or ex-cellent (.0.75) (15).

For the clinical study, unpaired t tests were used to compare demographic data between healthy volunteers and patients with CTS. Statistical evaluation of the US data from patients with CTS and volun-teers was complicated by the potential correlation of measurements in the two wrists of a single person. A logistic re-gression model with adjustment for clus-tering of wrists within patients was used to test differences in the AC/MN strain ratio and CSA between patients with CTS and control subjects. Optimal test thresh-olds for the AC/MN strain ratio and CSA were empirically defined based on re-ceiver operating characteristic (ROC) curves by choosing cutoffs that maxi-mized the sum of sensitivity and speci-ficity. To compute sensitivity, specificity, and 95% confidence intervals (CIs) for AC/MN strain ratio and CSA measure-ments to predict CTS, we used a gener-alized estimating equation approach to allow for potential correlation of wrist measurements within the same patient (16). The diagnostic performances of the AC/MN strain ratio and CSA were compared by using ROC analysis based on logistic models for each CSA and AC/MN strain ratio, which accounts for cor-relation within patients with the cluster option (17). A third logistic model was created for the combination of CSA and AC/MN strain ratio data; ROC analysis was used to compare the diagnostic ac-curacy of this new combined model with those of the independent CSA and AC/MN strain ratio models. The difference between areas under the ROC curves was compared by using the method de-scribed by Cleves (17) to accommodate clustered data.

All calculations were performed by using software (Stata, version 12.1; Sta-ta, College Station, Tex). P values less than .05 were considered to indicate a statistically significant difference.

Results

The preliminary examination consisted of examination of both hands in 10 partici-pants (three men [mean age, 60.7 years;

performing supramaximal percutaneous stimulation with a constant current stim-ulator and surface recordings were used. All NCVs were measured within 2 weeks of (either before or after) US examina-tion. Exclusion criteria for the patient group included the presence of a scar in the wrist from prior trauma or surgery, joint disorders such as rheumatoid ar-thritis, and a bifid MN at conventional B-mode US. Those who had a history of steroid injection into the carpal tunnel within 3 months of the study were also excluded. All patients completed a visual analog scale for pain (0–100 mm). Table 1 shows the baseline characteristics of the control subjects and the patients with CTS included in this study.

The CSA was measured three times by a radiologist trained in musculoskel-etal US (M.K., with 5 years of experi-ence), and the mean value was used for further analyses. The sonoelastographic examination was then performed by a second radiologist (A.S.K., with 5 years of experience in musculoskeletal sono-elastography). The B-mode US and so-noelastography examiners were blinded to the diagnosis of CTS, and both of them were not permitted to ask the volunteers or patients about their symp-toms. They were also blinded to the US measurements made by each other.

For the preliminary examination, the interobserver agreements for the AC/MN strain ratio were assessed by using the in-traclass correlation coefficient (ICC) cal-culated by using a two-way mixed-effects model to account for the random effect of patient hand and the fixed effect of the three observers. We report the ICC both for individual and averaged readings. The

prior trauma or surgery, joint disorders such as rheumatoid arthritis, a bifid MN at conventional B-mode US, and CTS symptoms. Three independent ex-aminers conducted this preliminary ex-amination. Observer A (M.K.) received training in musculoskeletal sonoelastog-raphy for 4 years. Observer B (A.S.K.) was a musculoskeletal radiologist with 5 years of sonoelastographic experience. Observer C (H.M.) received training in musculoskeletal sonoelastography for 3 years. All participants were examined by these three observers independently, and the observers were blinded to the medical history of the participants and the measurements assessed by the other observers. The interobserver data from this preliminary examination were not included in the clinical study.

For the clinical study, we used both a control group and a patient group. Exclusion criteria for the control group included the presence of a scar in the wrist from prior trauma or surgery, joint disorders such as rheumatoid arthritis, bifid MN at conventional B-mode US, and CTS symptoms. NCV testing was not performed in the control group. Both hands were evaluated for each control subject. To minimize observer bias, US and sonoelastographic exami-nations were performed for both hands in the patient group as well. However, for the study analysis, we included data only from symptomatic hands.

The diagnosis of CTS in our study was based on results of clinical and NCV testing (motor latencies .4.0 m/sec, sensory latencies .3.7 m/sec, am-plitudes ,20 mV, or conduction velocity ,50 m/sec) (1). Standard techniques of

Table 1

Baseline Demographic Data for Patients with CTS and Healthy Volunteers

Characteristic Patients with CTS (n = 31) Healthy Volunteers (n = 22) P Value

Age (y)* 62.0 6 13.3 60.9 6 13.2 .77Sex† .68 Female 28 (90) 19 (86) Male 3 (10) 3 (14)Visual analog scale for pain (0–100 mm)* 56.2 6 22.9 ... ...

* Data are means 6 standard deviations.† Data are numbers of patients or volunteers, with percentages in parentheses.



Logistic regression analysis demon-strated that AC/MN strain ratio and CSA were each significant independent predic-tors for the presence of CTS (P , .001). On the basis of logistic regression, the best predictive model combining AC/MN strain ratio with CSA for the diagnosis of CTS was as follows: 29.75 + 0.64 · CSA + 0.45 · AC/MN strain ratio, with a cutoff threshold of 0.55 to maximize the sum of sensitivity and specificity. The empirical threshold (0.55) correctly indicated 34 of 43 wrists with disease and 40 of 44 healthy wrists, providing a sensitivity of 81% (95% CI: 65%, 91%) with a spec-ificity of 91% (95% CI: 79%, 96%). As noted for strain ratio and CSA, the value for sensitivity computed by means of the generalized estimating equation method for clustered data differs slightly from the value that might be calculated directly from the correctly identified proportion. The area under the ROC curve for this combined model was 0.91 (95% CI: 0.85 0.97); compared with either CSA or strain ratio alone, the combined model provided significantly better predictive in-formation for the presence of CTS (P , .001). These results are shown in Table 2.

Discussion

Sonoelastography has enabled clini-cians to perform in situ evaluation of soft-tissue elasticity. In our study, we evaluated the elasticity of the MN by using a semiquantitative strain ratio.

difference in CSA between patients with CTS (13.7 mm2) and healthy vol-unteers (9.9 mm2; P , .001).

Diagnostic Accuracy of the AC/MN Strain Ratio

On the basis of empirical ROC analysis, the optimal test thresholds to maximize the sum of sensitivity and specificity for detection of CTS were defined by an AC/MN strain ratio cutoff of 4.3 and a CSA cutoff of 11 mm2. The AC/MN strain ra-tio correctly indicated 35 of 43 wrists with disease and 30 of 44 healthy wrists, providing a sensitivity of 82% (95% CI: 66%, 91%) at a specificity of 68% (95% CI: 53%, 80%), with an area under the ROC curve of 0.78 (95% CI: 0.69, 0.88). The CSA of the MN correctly indicated 35 of 43 wrists with disease and 33 of 44 healthy wrists, providing a sensitivity of 82% (95% CI: 66%, 91%) at a specific-ity of 75% (95% CI: 61%, 85%), with an area under the ROC curve of 0.85 (95% CI: 0.78, 0.93). The values for sensitiv-ity computed by means of the gener-alized estimating equation method for clustered data differ slightly from the values that might be calculated directly from the correctly identified propor-tions. Although the area under the ROC measured by means of the CSA was slightly higher than that measured by means of the AC/MN strain ratio, the difference was not statistically sig-nificant (P = .28).

age range, 50–80 years]; seven women [mean age, 60.9 years; age range, 39–81 years]). The control group consisted of both hands in 22 healthy volunteers (three men [mean age, 52.7 years; age range, 41–65 years]; 19 women [mean age, 62.2 years; age range, 40–88 years]). The patient group consisted of 43 hands in 31 patients with symp-tomatic CTS (three men [mean age, 69.0 years; age range, 46–88 years]; 28 women [mean age, 61.2 years; age range, 39–92 years]).

Interobserver Agreement for AC/MN Strain RatioOn the basis of the measurements re-corded by our three observers before the clinical study, the ICC for individual measurements by independent exam-iners in the same target wrist was 0.62 (95% CI: 0.38, 0.81). The ICC for av-erage measurements in the same target wrist was 0.83 (95% CI: 0.65, 0.93), suggesting that correlation is good for independent measurements and excel-lent for averaged measurements.

CSA and Elasticity of the MNRepresentative sonoelastographic im-ages of participants are shown in Fig-ure 1 and Figure 2. The mean AC/MN strain ratio in patients with CTS was 6.9, which was significantly higher than that in the healthy volunteers (4.1; P , .001). There was also a significant

Figure 1

Figure 1: Transverse images in 57-year-old healthy female volunteer. (a) Conventional B-mode US image shows the MN CSA corresponding to the circle with an area of 8 mm2. ∗ = Ulnar artery, FCR = flexor carpi radialis, P = pisiform bone, S = scaphoid bone. (b) Sonoelastographic image obtained at same level as a. The color represents the elasticity of the tissue within the region of interest, whose scale ranged from red for components with the greatest strain (softest components) to blue for those with no strain (hardest components). The strain ratio of the AC (B) to the MN (A) was 2.9.



be helpful in diagnosing CTS in combina-tion with CSA.

Our study had several limitations. We measured the elasticity of the MN at only one point (the scaphoid-pisiform bone level) on a transverse plane. The MN is a cylindric, symmetric structure. Given this geometry, the result of the sonoelastographic measurements may be variable related to the anisotropy at the boundary of the object, especially on the transverse plane (25). The ICC of the semiquantitative strain ratio has al-ready been investigated in normal Achil-les tendons in the transverse and longi-tudinal planes (0.41–0.78); the ICC was higher for the longitudinal than for the transverse plane measurements (26). To allow for accurate sonoelastographic evaluation, the transducer should always be held perpendicular to the MN. The MN runs parallel to the skin and flexor tendons with slight flexion of the wrist at that level (27). Our study results showed diagnostic accuracy for CTS on the basis of clinical symptoms and NCV testing results as a reference standard. NCV has been widely used in the diagnosis of CTS, whereas the false-negative rate for NCV testing has been reported to be between 16% and 34% (28). It is possi-ble that we did not include patients with CTS whose findings were negative at NCV testing (ie, they had false-negative findings).

nerve that had undergone chronic com-pression experienced epineurial fibrosis and thickening (20). The stiffer MN in the patient with CTS might reflect these degenerative changes in the MN.

Misdiagnosis has been identified as one of the most common causes of treat-ment failure for CTS (21). Prior studies in which investigators have evaluated the diagnostic accuracy of CSA as a single test, with clinical provocation and NCV tests as the reference standard, had 82%–94% sensitivities and 65%–97% specificities (7,22–24). In our study, the combined application of AC/MN strain ratio and CSA provided improved ac-curacy in the diagnosis of CTS with a sensitivity of 81% (34 of 43) and a spec-ificity of 91% (40 of 44). Therefore, de-termining the elasticity of the MN could

The interobserver reproducibility of the AC/MN strain ratio was found to be good to excellent in our preliminar-ily investigation. On the basis of our clinical study results, the MN in pa-tients with CTS demonstrates a higher AC/MN strain ratio compared with that of healthy volunteers, as well as a greater CSA. The combination of these two features provided significantly bet-ter diagnostic accuracy for CTS than did measurement of either AC/MN strain ratio or MN CSA alone.

The pathophysiology of CTS is be-lieved to be a combination of increased intracarpal tunnel pressure and ischemic injury in the MN, which lead to focal de-myelination and axonal degeneration with the fibrotic response (18,19). There are four documented cases in which a human

Figure 2

Figure 2: Transverse images obtained in 64-year-old woman with CTS. (a) Conventional B-mode US image shows the MN CSA corresponding to the circle with an area of 18 mm2. ∗ = Ulnar artery, FCR = flexor carpi radialis, P = pisiform bone, S = scaphoid bone. (b) Sonoelastographic image obtained at the same level as the B-mode image. The strain ratio of the AC (B) to the MN (A) was 7.6.

Table 2

Diagnostic Accuracy of AC/MN Strain Ratio, CSA, and Combined Logistic Model for CTS

Diagnostic Test Sensitivity (%) Specificity (%)Area under the ROC Curve

AC/MN strain ratio 4.3 82 (35/43) [66%, 91%] 68 (30/44) [53%, 80%] 0.78 [0.69, 0.88]CSA 11 mm2 82 (35/43) [66%, 91%] 75 (33/44) [61%, 85%] 0.85 [0.78, 0.93]Combined logistic model 0.55* 81 (34/43) [65%, 91%] 91 (40/44) [79%, 96%] 0.91 [0.85, 0.97]†

Note.—Data in parentheses are numbers of wrists used to calculate the percentages; data in brackets are 95% CIs.

* Combined logistic model = 29.75 + 0.64 · CSA + 0.45 · AC/MN strain ratio.† P , .001.



Although we evaluated compression elastography, several additional quan-titative sonoelastographic methods are commercially available, including shear wave elastography, transient elastog-raphy, and acoustic force elastography (29–31). These three techniques might not be suitable for our purpose because they are available only for regional mea-surement with limited depth or cannot depict anatomic images (29–31). In our study, we focused on sonoelastogra-phy’s usefulness and findings for CTS. A more accurate method for CTS diagno-sis by using a change in CSA equal to or greater than 2 mm2 between proximal and distal levels has been reported (6).

Sonoelastographic equipment can display real-time conventional B-mode and sonoelastographic images simulta-neously; it can also allow for US-guid-ed injection once images are obtained. Therefore, a technique combining both elasticity and CSA measures of the MN as diagnostic indicator may be prefera-ble for use in clinical practice. We con-clude that the MN in patients with CTS is stiffer than that in healthy subjects and that sonoelastography provides sig-nificant improvement in the diagnostic accuracy of the US assessment of CTS. Further studies are necessary to see how sonoelastography complements CSA measurements for evaluation of CTS.

Disclosures of Conflicts of Interest: H.M. No rel-evant conflicts of interest to disclose. E.J.H. No relevant conflicts of interest to disclose. M.K. No relevant conflicts of interest to disclose. M.G. No relevant conflicts of interest to disclose. R.A. No relevant conflicts of interest to disclose. R.B. No relevant conflicts of interest to disclose. G.M.F. No relevant conflicts of interest to disclose. W.R.J. No relevant conflicts of interest to disclose. A.S.K. No relevant conflicts of interest to disclose.

References 1. Keith MW, Masear V, Chung KC, et al.

American Academy of Orthopaedic Surgeons Clinical Practice Guideline on diagnosis of carpal tunnel syndrome. J Bone Joint Surg Am 2009;91(10):2478–2479.

2. Buchberger W. Radiologic imaging of the car-pal tunnel. Eur J Radiol 1997;25(2):112–117.

3. Wong SM, Griffith JF, Hui AC, Tang A, Wong KS. Discriminatory sonographic criteria for

the diagnosis of carpal tunnel syndrome. Ar-thritis Rheum 2002;46(7):1914–1921.

4. Altinok T, Baysal O, Karakas HM, et al. Ul-trasonographic assessment of mild and mod-erate idiopathic carpal tunnel syndrome. Clin Radiol 2004;59(10):916–925.

5. Ziswiler HR, Reichenbach S, Vögelin E, Bachmann LM, Villiger PM, Jüni P. Diagnos-tic value of sonography in patients with sus-pected carpal tunnel syndrome: a prospective study. Arthritis Rheum 2005;52(1):304–311.

6. Klauser AS, Halpern EJ, De Zordo T, et al. Carpal tunnel syndrome assessment with US: value of additional cross-sectional area measurements of the median nerve in pa-tients versus healthy volunteers. Radiology 2009;250(1):171–177.

7. Mohammadi A, Afshar A, Etemadi A, Masoudi S, Baghizadeh A. Diagnostic value of cross-sectional area of median nerve in grading severity of carpal tunnel syndrome. Arch Iran Med 2010;13(6):516–521.

8. Tai TW, Wu CY, Su FC, Chern TC, Jou IM. Ultrasonography for diagnosing carpal tunnel syndrome: a meta-analysis of diag-nostic test accuracy. Ultrasound Med Biol 2012;38(7):1121–1128.

9. Onur MR, Poyraz AK, Ucak EE, Bozgeyik Z, Özercan IH, Ogur E. Semiquantitative strain elastography of liver masses. J Ultrasound Med 2012;31(7):1061–1067.

10. Fischer T, Peisker U, Fiedor S, et al. Signif-icant differentiation of focal breast lesions: raw data-based calculation of strain ratio. Ultraschall Med 2012;33(4):372–379.

11. Itokawa F, Itoi T, Sofuni A, et al. EUS elas-tography combined with the strain ratio of tissue elasticity for diagnosis of solid pancre-atic masses. J Gastroenterol 2011;46(6):843–853.

12. Zhang Y, Tang J, Li YM, et al. Differentiation of prostate cancer from benign lesions using strain index of transrectal real-time tissue elastography. Eur J Radiol 2012;81(5):857–862.

13. Ning CP, Jiang SQ, Zhang T, Sun LT, Liu YJ, Tian JW. The value of strain ratio in differen-tial diagnosis of thyroid solid nodules. Eur J Radiol 2012;81(2):286–291.

14. Havre RF, Elde E, Gilja OH, et al. Freehand real-time elastography: impact of scanning parameters on image quality and in vitro intra- and interobserver validations. Ultra-sound Med Biol 2008;34(10):1638–1650.

15. Fleiss JL. The design and analysis of clinical experiments. New York, NY: Wiley, 1986.

16. Genders TS, Spronk S, Stijnen T, Steyer-berg EW, Lesaffre E, Hunink MG. Methods for calculating sensitivity and specificity

of clustered data: a tutorial. Radiology 2012;265(3):910–916.

17. Cleves MA. From the help desk: comparing areas under receiver operating characteris-tic curves from two or more probit or logit models. Stata J 2002;2(3):301–313.

18. Ibrahim I, Khan WS, Goddard N, Smitham P. Carpal tunnel syndrome: a review of the re-cent literature. Open Orthop J 2012;6:69–76.

19. Rempel D, Dahlin L, Lundborg G. Patho-physiology of nerve compression syndromes: response of peripheral nerves to loading. J Bone Joint Surg Am 1999;81(11):1600–1610.

20. Mackinnon SE, Dellon AL, Hudson AR, Hunter DA. Chronic human nerve compres-sion: a histological assessment. Neuropathol Appl Neurobiol 1986;12(6):547–565.

21. Hunt TR, Osterman AL. Complications of the treatment of carpal tunnel syndrome. Hand Clin 1994;10(1):63–71.

22. Wong SM, Griffith JF, Hui AC, Lo SK, Fu M, Wong KS. Carpal tunnel syndrome: di-agnostic usefulness of sonography. Radiology 2004;232(1):93–99.

23. Yesildag A, Kutluhan S, Sengul N, et al. The role of ultrasonographic measurements of the median nerve in the diagnosis of carpal tunnel syndrome. Clin Radiol 2004;59(10):910–915.

24. Duncan I, Sullivan P, Lomas F. Sonography in the diagnosis of carpal tunnel syndrome. AJR Am J Roentgenol 1999;173(3):681–684.

25. Thitaikumar A, Krouskop TA, Garra BS, Ophir J. Visualization of bonding at an inclu-sion boundary using axial-shear strain elas-tography: a feasibility study. Phys Med Biol 2007;52(9):2615–2633.

26. Drakonaki EE, Allen GM, Wilson DJ. Real-time ultrasound elastography of the normal Achilles tendon: reproducibility and pattern description. Clin Radiol 2009;64(12):1196–1202.

27. Bianchi S, Montet X, Martinoli C, Bonvin F, Fasel J. High-resolution sonography of com-pressive neuropathies of the wrist. J Clin Ul-trasound 2004;32(9):451–461.

28. Witt JC, Hentz JG, Stevens JC. Carpal tun-nel syndrome with normal nerve conduction studies. Muscle Nerve 2004;29(4):515–522.

29. Parker KJ, Fu D, Graceswki SM, Yeung F, Levinson SF. Vibration sonoelastography and the detectability of lesions. Ultrasound Med Biol 1998;24(9):1437–1447.

30. Sandrin L, Catheline S, Tanter M, Hennequin X, Fink M. Time-resolved pulsed elastogra-phy with ultrafast ultrasonic imaging. Ultra-son Imaging 1999;21(4):259–272.

31. Walker WF. Internal deformation of a uni-form elastic solid by acoustic radiation force. J Acoust Soc Am 1999;105(4):2508–2518.

Carpal Tunnel Syndrome: Diagnosis by Means of Median Nerve Elasticity—Improved Diagnostic Accuracy...

Documents

Transcript of Carpal Tunnel Syndrome: Diagnosis by Means of Median Nerve Elasticity—Improved Diagnostic Accuracy...