JMIR Publications

ABSTRACT

Background:

Carpal tunnel syndrome (CTS) is a medical condition caused by compression of the median nerve in the carpal tunnel due to aging or overuse of the hand. The symptoms include numbness of the fingers and atrophy of the thenar muscle. Thenar atrophy recovers slowly postoperatively; therefore, early diagnosis and surgery are important. While physical examinations and nerve conduction studies are used to diagnose CTS, problems with the diagnostic ability and equipment, respectively, exist. Despite research on the application for screening CTS using a tablet and machine learning, problems with the usage rate of tablets and data collection for machine learning remain.

Objective:

To make data collection for machine learning easier and more available, we developed a screening application for CTS using a smartphone and an anomaly detection algorithm, and aimed to examine our system as a useful screening tool for CTS.

Methods:

In total, 36 CTS hands and 27 non-CTS hands were recruited. Participants controlled the character in our application using their thumbs. We recorded the position of the thumbs and time, and generated screening models that classify CTS and non-CTS using anomaly detection and an autoencoder and calculated the sensitivity, specificity, and area under the curve (AUC).

Results:

CTS and non-CTS participants were classified with 93% sensitivity, 69% specificity, and 0.86 AUC. When dividing the data by direction, the model with data in the same direction as the thumb opposition had the highest AUC of 0.99, 92% sensitivity, and 100% specificity.

Conclusions:

Our application could reveal the difficulty of thumb opposition for CTS patients and screen for CTS with high sensitivity and specificity. The application is highly accessible because of the use of smartphones and can easily enhance machine learning using anomaly detection.