Carpal Tunnel Syndrome Automated Diagnosis: A Motor vs. Sensory Nerve Conduction-Based Approach

Abstract

The objective of this study was to evaluate the effectiveness of machine learning classification techniques applied to nerve conduction studies (NCS) of motor and sensory signals for the automatic diagnosis of carpal tunnel syndrome (CTS). Two methodologies were tested. In the first methodology, motor signals recorded from the patients' median nerve were transformed into time-frequency spectrograms using the short-time Fourier transform (STFT). These spectrograms were then used as input to a deep two-dimensional convolutional neural network (CONV2D) for classification into two categories: patients and controls. In the second methodology, sensory signals from the patients' median and ulnar nerves were subjected to multilevel wavelet decomposition (MWD), and statistical and non-statistical features were extracted from the decomposed signals. These features were utilized to train and test classifiers. The classification target was set to three categories: normal subjects (controls), patients with mild CTS, and patients with moderate to severe CTS based on conventional electrodiagnosis results. The results of the classification analysis demonstrated that both methodologies surpassed previous attempts at automatic CTS diagnosis. The classification models utilizing the motor signals transformed into time-frequency spectrograms exhibited excellent performance, with average accuracy of 94%. Similarly, the classifiers based on the sensory signals and the extracted features from multilevel wavelet decomposition showed significant accuracy in distinguishing between controls, patients with mild CTS, and patients with moderate to severe CTS, with accuracy of 97.1%. The findings highlight the efficacy of incorporating machine learning algorithms into the diagnostic processes of NCS, providing a valuable tool for clinicians in the diagnosis and management of neuropathies such as CTS.

Keywords: carpal tunnel syndrome; deep learning; machine learning; multilevel wavelet decomposition; nerve conduction studies; short-time Fourier transform; spectrogram.

Figure 1

Depiction of two sensory signals from a patient with severe symptoms of CTS. The orange-colored signal was extracted from an ulnar nerve, while the blue-colored signal was extracted from a median nerve.

Figure 2

The waveform of a standard motor signal from the median nerve, contrasted with that of a motor signal affected by CTS.

Figure 3

Spectrograms of two participants belonging to each class of the classification problem. The arrow indicates the transformation of the signal when applying STFT to it, while the different colors of the spectrogram indicate the amplitude of a particular frequency at a particular time.

Figure 4

The architecture of the CONV2D classification model.

Figure 5

Accuracy and loss value curves of the proposed COVN2D model with different values of learning rate.

Figure 6

Multilevel discrete wavelet decomposition of a signal with a depth value of 5.

Figure 7

The confusion matrix of the two-class prediction.

Figure 8

The normalized confusion matrix of the best result achieved, using the random forest (RF) classifier.