Related Experiment Video
Updated: Oct 12, 2025
06:40
Screening of Axonal Degeneration in Carpal Tunnel Syndrome Using Ultrasonography and Nerve Conduction Studies
Published on: January 11, 2019
11.8K
Automatic Electrodiagnosis of Carpal Tunnel Syndrome Using Machine Learning
Konstantinos I Tsamis1,2, Prokopis Kontogiannis3, Ioannis Gourgiotis1
1Department of Neurology, University Hospital of Ioannina, 45110 Ioannina, Greece.
Bioengineering (Basel, Switzerland)
|November 25, 2021
Summary
This study demonstrates that machine learning can automatically detect carpal tunnel syndrome with high accuracy using electrodiagnostic features. This automated approach aids in clinical decision-making and reduces potential human error in diagnosis.
Area of Science:
- Neurology
- Biomedical Engineering
- Machine Learning in Medicine
Background:
- Carpal tunnel syndrome (CTS) diagnosis relies heavily on nerve conduction studies (NCS).
- There is ongoing debate regarding the necessity and optimal use of NCS in CTS management.
- Accurate and efficient diagnostic tools for CTS are crucial for timely patient care.
Purpose of the Study:
- To investigate the feasibility of automatically detecting median nerve mononeuropathy and diagnosing CTS.
- To utilize electrodiagnostic features, including novel physiological and mathematical metrics, for automated diagnosis.
- To assess the accuracy of machine learning-based electrodiagnosis compared to physician diagnoses.
Main Methods:
- Prospective study involving 38 volunteers.
- Application of machine learning techniques to combine electrodiagnostic features for diagnosis.
- Inclusion of both common clinical electrodiagnostic criteria and novel features.
Main Results:
- Automatic electrodiagnosis achieved 95% accuracy against standard neurophysiological diagnosis by physicians.
- The automated system showed 89% accuracy compared to clinical diagnosis.
- Novel electrodiagnostic features, alongside common ones, improved diagnostic accuracy.
Conclusions:
- Automated detection of carpal tunnel syndrome using machine learning is feasible and accurate.
- This technology can assist in clinical decision-making and potentially minimize human error.
- The integration of novel electrodiagnostic features enhances the reliability of automated CTS detection.
