March 15, 2021
Japan Science and Technology Agency (JST)
Tokyo Medical and Dental University
Keio University
Under the JST Strategic Basic Research Programs, a research group led by Lecturer Koji Fujita of the Graduate School of Medical and Dental Sciences at Tokyo Medical and Dental University and Associate Professor Yuta Sugiura of the Faculty of Science and Technology at Keio University has developed a simple method for screening for carpal tunnel syndrome by combining machine learning using an anomaly detection method with the analysis of thumb movement via a smartphone app.
Carpal tunnel syndrome, common in middle-aged and older women, is a condition caused by pressure on a nerve in the wrist, leading to numbness in the hands and difficulty moving the fingers. While a nerve conduction velocity test can provide an accurate diagnosis, it is not widely available as it requires expensive equipment and specialized skills. There is a need for a simple screening tool that can be used without specialized knowledge or skills.
The research group focused on the deterioration of thumb movement as the disease progresses and analyzed its characteristics. They developed a smartphone game app played with the thumb, acquired data on the thumb's trajectory during gameplay, and created a program that uses machine learning to estimate the presence or absence of the disease. Users can be screened for potential carpal tunnel syndrome simply by playing a short game for about 30 seconds to one minute. By using an anomaly detection method, the team efficiently built an estimation model from the data of just 12 healthy subjects, even without a large dataset from patients with the condition.
The developed tool will enable screening for potential carpal tunnel syndrome even in environments without medical specialists, such as at home or in public health centers. In the future, the group aims to develop a system that encourages users with suspected cases to consult a specialist, thereby helping to prevent the condition from becoming severe. They believe that by preventing the disability and social loss associated with the progression of this disease, which is common in women, this work can contribute to a society where women can thrive.
The results of this research will be published online in the international scientific journal "JMIR mHealth and uHealth" on March 14, 2021 (US Eastern Daylight Time).
For the full press release, please see below.