腕隧道症候群是一種常見的手部疾病，主要是由於手腕神經受到壓迫產生不適感。在過去研究中已發現多個與該疾病相關的量化特徵，並藉由病理學得到合理的解釋，如神經面積在患病過程中會因為神經腫脹造成面積變大的情況。但許多相關的參數主要多集中於與疾病嚴重程度的相關性，較少關注於治療後復原情況的評估，並且在每一個文獻中都會提出所主張的判斷依據，對此可能會產生標準不一的情況。因此，本研究希望藉由引入人工智慧分類器(K-nearest neighbor、naïve Bayes、support vector machine、decision tree)用以預測及分類腕隧道的患病與否，並且針對手術治療的患者進行量化分析。本研究使用商用超音波儀器及12 MHz探頭針對三十位的健康受試者及二十二位患者進行腕部超音波影像掃描，並有十二位患者提供術後超音波影像。該資料經過特徵提取將用於分類器訓練及測試使用。結果顯示四種分類器可達約75-80% 的準確度，其中support vector machine 可達83%的敏感度，naïve Bayes可達到87%的特異度。在術後資料分析結果中發現，與疾病嚴重程度相關之量化參數在術後資料群與其他群具有顯著差異，而經過該參數訓練之分類器亦能辨別在術後群體間的差異，約有半數資料經過分類器預測已為健康狀態。本研究所提出使用分類器進行腕隧道症候群患病評估可做後續電腦輔助診斷之開創。

Carpal tunnel syndrome (CTS) is a common wrist disease. The cause is mainly due to the compression of wrist median nerve and results in discomfort chronically. In the past research, many quantitative parameters have been proven to be related with CTS according to pathological reason. For example, the nerve cross-sectional area may increase due the inflammation of median nerve. Most of these parameters address with the severity of CTS. Almost none of them have been studied to evaluate the recovery situation after therapy. There is another problem that exists in previous studies. For each study, the evaluation may differ significantly, which cannot be a common gold standard. Therefore, the purpose of this study is to introduce the artificial intelligence classifiers including k-nearest neighbor, naïve Bayes, support vector machine, and decision trees to predict and evaluate whether a subject is suffering with CTS or not. In addition, post-surgery patients data are also collected to find the relationship between recovery status and the quantitative parameters. A commercial ultrasound system is used with a 12 MHz probe to record the wrist ultrasound images. 30 health subjects and 22 CTS patients are enrolled. 12 of the CTS patients also received the ultrasound scans after surgery. The ultrasound images are then utilized to extract features for classifier training and testing. The results show that the classification of these four classifier can reach 74-80% accuracy. Support vector machine reaches the highest sensitivity of 83%. Naïve Bayes classifier reaches the highest specificity of 87%. The post-surgery patient data yielded significant difference of extracted parameters from the data of health subjects and CTS patients before the operation. The results show classifiers are able to distinguished the difference within post-surgical data. These classifiers used in this study can be used for computer-aided systems to assist the evaluation of CTS.

[1] K. K. Shung, Diagnostic ultrasound: Imaging and blood flow measurements. CRC press, 2015.
[2] H. Strunk et al., "Clinical use of High-Intensity Focused Ultrasound (HIFU) for tumor and pain reduction in advanced pancreatic cancer," in RöFo-Fortschritte auf dem Gebiet der Röntgenstrahlen und der bildgebenden Verfahren, 2016, vol. 188, no. 07, pp. 662-670: © Georg Thieme Verlag KG.
[3] R. J. van den Bijgaart, D. C. Eikelenboom, M. Hoogenboom, J. J. Fütterer, M. H. den Brok, and G. J. Adema, "Thermal and mechanical high-intensity focused ultrasound: perspectives on tumor ablation, immune effects and combination strategies," Cancer Immunology, Immunotherapy, vol. 66, no. 2, pp. 247-258, 2017.
[4] M. Kassinopoulos, E. Bousi, I. Zouvani, and C. Pitris, "Correlation of the derivative as a robust estimator of scatterer size in optical coherence tomography (OCT)," Biomedical optics express, vol. 8, no. 3, pp. 1598-1606, 2017.
[5] D. H. Solomon, J. N. Katz, R. Bohn, H. Mogun, and J. Avorn, "Nonoccupational risk factors for carpal tunnel syndrome," Journal of general internal medicine, vol. 14, no. 5, pp. 310-314, 1999.
[6] S. S. Karne and N. S. Bhalerao, "Carpal tunnel syndrome in hypothyroidism," Journal of clinical and diagnostic research: JCDR, vol. 10, no. 2, p. OC36, 2016.
[7] I. Ibrahim, W. Khan, N. Goddard, and P. Smitham, "Suppl 1: Carpal Tunnel Syndrome: A Review of the Recent Literature," The open orthopaedics journal, vol. 6, p. 69, 2012.
[8] N. Lam and A. Thurston, "Association of obesity, gender, age and occupation with carpal tunnel syndrome," ANZ Journal of Surgery, vol. 68, no. 3, pp. 190-193, 1998.
[9] I. Atroshi, C. Gummesson, R. Johnsson, E. Ornstein, J. Ranstam, and I. Rosén, "Prevalence of carpal tunnel syndrome in a general population," Jama, vol. 282, no. 2, pp. 153-158, 1999.
[10] S. H. Kuschner, E. Ebramzadeh, D. Johnson, W. W. Brien, and R. Sherman, "Tinel's sign and Phalen's test in carpal tunnel syndrome," Orthopedics, vol. 15, no. 11, pp. 1297-1302, 1992.
[11] S. TINEL’S, "Tinel’s sign and Phalen’s maneuver: physical signs of carpal tunnel syndrome," Hospital Physician, p. 39, 2000.
[12] P. Seror, "Phalen’s test in the diagnosis of carpal tunnel syndrome," Journal of Hand Surgery, vol. 13, no. 4, pp. 383-385, 1988.
[13] H. Gelmers, "The significance of Tinel's sign in the diagnosis of carpal tunnel syndrome," Acta neurochirurgica, vol. 49, no. 3-4, pp. 255-258, 1979.
[14] J.-i. Iida, H. Hirabayashi, H. Nakase, and T. Sakaki, "Carpal tunnel syndrome: electrophysiological grading and surgical results by minimum incision open carpal tunnel release," Neurologia medico-chirurgica, vol. 48, no. 12, pp. 554-559, 2008.
[15] L. Heller, H. Ring, H. Costeff, and P. Solzi, "Evaluation of Tinel’s and Phalen’s signs in diagnosis of the carpal tunnel syndrome," European neurology, vol. 25, no. 1, pp. 40-42, 1986.
[16] H. L. Lew, E. S. Date, S. S. Pan, P. Wu, P. F. Ware, and W. S. Kingery, "Sensitivity, specificity, and variability of nerve conduction velocity measurements in carpal tunnel syndrome," Archives of physical medicine and rehabilitation, vol. 86, no. 1, pp. 12-16, 2005.
[17] E. Silvestri, C. Martinoli, L. E. Derchi, M. Bertolotto, M. Chiaramondia, and I. Rosenberg, "Echotexture of peripheral nerves: correlation between US and histologic findings and criteria to differentiate tendons," Radiology, vol. 197, no. 1, pp. 291-296, 1995.
[18] L. Sarria, T. Cabada, R. Cozcolluela, T. Martinez-Berganza, and S. Garcia, "Carpal tunnel syndrome: usefulness of sonography," European radiology, vol. 10, no. 12, pp. 1920-1925, 2000.
[19] P. Molitor, "A diagnostic test for carpal tunnel syndrome using ultrasound," Journal of Hand Surgery, vol. 13, no. 1, pp. 40-41, 1988.
[20] L. Padua et al., "Carpal tunnel syndrome: ultrasound, neurophysiology, clinical and patient-oriented assessment," Clinical Neurophysiology, vol. 119, no. 9, pp. 2064-2069, 2008.
[21] L. Moran, M. Perez, A. Esteban, J. Bellon, B. Arranz, and M. del Cerro, "Sonographic measurement of cross‐sectional area of the median nerve in the diagnosis of carpal tunnel syndrome: Correlation with nerve conduction studies," Journal of Clinical Ultrasound, vol. 37, no. 3, pp. 125-131, 2009.
[22] A. M. Ettema, P. C. Amadio, C. Zhao, L. E. Wold, and K.-N. An, "A histological and immunohistochemical study of the subsynovial connective tissue in idiopathic carpal tunnel syndrome," JBJS, vol. 86, no. 7, pp. 1458-1466, 2004.
[23] J. W. H. Korstanje et al., "Ultrasonographic assessment of longitudinal median nerve and hand flexor tendon dynamics in carpal tunnel syndrome," Muscle & nerve, vol. 45, no. 5, pp. 721-729, 2012.
[24] Y.-H. Lin, M.-Y. Hsieh, F.-C. Su, and S.-H. Wang, "Assessment of the kinetic trajectory of the median nerve in the wrist by high-frequency ultrasound," Sensors, vol. 14, no. 5, pp. 7738-7752, 2014.
[25] A. Filius et al., "Multidimensional ultrasound imaging of the wrist: Changes of shape and displacement of the median nerve and tendons in carpal tunnel syndrome," Journal of Orthopaedic Research, vol. 33, no. 9, pp. 1332-1340, 2015.
[26] F. C. Arnett et al., "The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis," Arthritis & Rheumatology, vol. 31, no. 3, pp. 315-324, 1988.
[27] I. Androutsopoulos, G. Paliouras, V. Karkaletsis, G. Sakkis, C. D. Spyropoulos, and P. Stamatopoulos, "Learning to filter spam e-mail: A comparison of a naive bayesian and a memory-based approach," arXiv preprint cs/0009009, 2000.
[28] S. B. Kotsiantis, I. Zaharakis, and P. Pintelas, "Supervised machine learning: A review of classification techniques," Emerging artificial intelligence applications in computer engineering, vol. 160, pp. 3-24, 2007.
[29] F. Lotte, M. Congedo, A. Lécuyer, F. Lamarche, and B. Arnaldi, "A review of classification algorithms for EEG-based brain–computer interfaces," Journal of neural engineering, vol. 4, no. 2, p. R1, 2007.
[30] P. J. García-Laencina, J.-L. Sancho-Gómez, and A. R. Figueiras-Vidal, "Pattern classification with missing data: a review," Neural Computing and Applications, vol. 19, no. 2, pp. 263-282, 2010.
[31] R. Ranawana and V. Palade, "Multi-classifier systems: Review and a roadmap for developers," International Journal of Hybrid Intelligent Systems, vol. 3, no. 1, pp. 35-61, 2006.
[32] J. Fulcher, "Computational intelligence: an introduction," in Computational intelligence: a compendium: Springer, 2008, pp. 3-78.
[33] A. Leclerc et al., "Carpal tunnel syndrome and work organisation in repetitive work: a cross sectional study in France. Study Group on Repetitive Work," Occupational and environmental medicine, vol. 55, no. 3, pp. 180-187, 1998.
[34] B. A. Silverstein, L. J. Fine, and T. J. Armstrong, "Occupational factors and carpal tunnel syndrome," American journal of industrial medicine, vol. 11, no. 3, pp. 343-358, 1987.
[35] R. M. Szabo, "Carpal tunnel syndrome as a repetitive motion disorder," Clinical orthopaedics and related research, no. 351, pp. 78-89, 1998.
[36] T. Cover and P. Hart, "Nearest neighbor pattern classification," IEEE transactions on information theory, vol. 13, no. 1, pp. 21-27, 1967.
[37] D. Kumar, "Performance analysis of various data mining algorithms: A review," methods, vol. 32, no. 6, pp. 9-15, 1995.
[38] D. D. Lewis, "Naive (Bayes) at forty: The independence assumption in information retrieval," in European conference on machine learning, 1998, pp. 4-15: Springer.
[39] M. A. Hearst, S. T. Dumais, E. Osuna, J. Platt, and B. Scholkopf, "Support vector machines," IEEE Intelligent Systems and their applications, vol. 13, no. 4, pp. 18-28, 1998.
[40] S. R. Gunn, "Support vector machines for classification and regression," ISIS technical report, vol. 14, no. 1, pp. 5-16, 1998.
[41] J. A. Suykens and J. Vandewalle, "Least squares support vector machine classifiers," Neural processing letters, vol. 9, no. 3, pp. 293-300, 1999.
[42] S.-i. Amari and S. Wu, "Improving support vector machine classifiers by modifying kernel functions," Neural Networks, vol. 12, no. 6, pp. 783-789, 1999.
[43] J. Weston and C. Watkins, "Support vector machines for multi-class pattern recognition," in Esann, 1999, vol. 99, pp. 219-224.
[44] Y. Liu and Y. F. Zheng, "One-against-all multi-class SVM classification using reliability measures," Neural Networks, vol. 2, pp. 849-854, 2005.
[45] S. R. Safavian and D. Landgrebe, "A survey of decision tree classifier methodology," IEEE transactions on systems, man, and cybernetics, vol. 21, no. 3, pp. 660-674, 1991.
[46] M. Kayri and İ. Kayri, "The Comparison of Gini and Twoing Algorithms in Terms of Predictive Ability and Misclassification Cost in Data Mining: An Empirical Study," databases, vol. 3, p. 5.
[47] C.-H. Lin, "Assessment of Kinetic Trajectory of the Median Nerve from Wrist Ultrasound Images in Patients with Carpal Tunnel Syndrome by Motion Tracking Technique," 2017.
[48] L.-K. Kuo, "Assessment of Kinetic Trajectory of the Median Nerve from Wrist Ultrasound Images Using Adaptive Baysian Speckle Tracking Technique," 2014.
[49] A. S. Klauser et al., "Carpal tunnel syndrome assessment with US: value of additional cross-sectional area measurements of the median nerve in patients versus healthy volunteers," Radiology, vol. 250, no. 1, pp. 171-177, 2009.
[50] A. S. Klauser et al., "Sonographic cross-sectional area measurement in carpal tunnel syndrome patients: can delta and ratio calculations predict severity compared to nerve conduction studies?," European radiology, vol. 25, no. 8, pp. 2419-2427, 2015.
[51] A. L. Moreira and L. D. Hobson-Webb, "T111. Median nerve area in diabetic carpal tunnel syndrome patients: a retrospective review of 1030 cases," Clinical Neurophysiology, vol. 129, p. e45, 2018.