Automatic Electrodiagnosis of Carpal Tunnel Syndrome Using Machine Learning

Abstract

Recent literature has revealed a long discussion about the importance and necessity of nerve conduction studies in carpal tunnel syndrome management. The purpose of this study was to investigate the possibility of automatic detection, based on electrodiagnostic features, for the median nerve mononeuropathy and decision making about carpal tunnel syndrome. The study included 38 volunteers, examined prospectively. The purpose was to investigate the possibility of automatically detecting the median nerve mononeuropathy based on common electrodiagnostic criteria, used in everyday clinical practice, as well as new features selected based on physiology and mathematics. Machine learning techniques were used to combine the examined characteristics for a stable and accurate diagnosis. Automatic electrodiagnosis reached an accuracy of 95% compared to the standard neurophysiological diagnosis of the physicians with nerve conduction studies and 89% compared to the clinical diagnosis. The results show that the automatic detection of carpal tunnel syndrome is possible and can be employed in decision making, excluding human error. It is also shown that the novel features investigated can be used for the detection of the syndrome, complementary to the commonly used ones, increasing the accuracy of the method.

Added Value of Morphological and Functional Magnetic Resonance Neurography in Assessment of Carpal Tunnel Syndrome in Correlation with Nerve Conduction Studies

“…We reported the maximum FR regardless of the site of flattening, yet that was at the level of proximal carpal tunnel at most of the cases. On the contrary, another study, showed significant difference in FR at the proximal carpal tunnel (2.2± 0.52 in the control group & 2.7 ± 0.9 in the patient group, p value < 0.0001) with 72.5 % & 65.6 % specificity using a of cut off value >2.4 (3) . With the use of the same cut off value, our study showed 60 % sensitivity & 90 % specificity (p value =0.091).…”

“…These values are less than that suggested by other studies. For example, in another study, the sensitivity was 85 % with 100 % specificity using >15 mm 2 at pisiform bone as cut off value (3) . In another example the cut off value of CSA at pisiform bone was 14.4 mm 2 in another study, with sensitivity = 90 % (9) .…”

“…For example, a study found that SIR was significantly different between the two groups (1.0 ± 0.30 in control group & 1.30 ± 0.30 in patient group, p value < 0.0001) and with the use of 1.05 as cut off value, the sensitivity, specificity & accuracy for diagnosis of CTS were 88.4, 68.8, and 82.2%, respectively (3) . Also, in another study, the use of SIR > 2.4 as cut off value resulted in 70 % sensitivity and 66.7 % specificity (10) .…”

“…Many controversies have arisen regarding the necessity of nerve conduction studies (NCS) as confirmatory tests or tools contributing to patient management. These controversies emanate mainly by the rise of the time and the cost needed for the diagnosis of CTS with the use of NCS, the discomfort caused to the patients by nerve stimulation, and the high rate of false positive and false negative results of NCS along with the ease and the accuracy of clinical diagnosis in many cases with symptoms and signs typical of CTS (2) . When clinical assessment or NCS are somewhat equivocal regarding a diagnosis of CTS, hand surgeons may have hesitancy in undertaking surgery.…”

“…At this situation, ultrasound or MRI may be helpful in supporting a diagnosis of CTS. MRI has advantages over ultrasound in being less operator-dependent, allowing clearer delineation of the carpal tunnel contents, and enabling the entire median nerve to be measured (3) . MRI of peripheral nerves, also referred to as MR neurography (MRN), is increasingly being used in clinical routine because of advances in MRI hardware and the development of new imaging techniques.…”