Researchers at the Indian Institute of Science (IISc) in Bengaluru have developed an artificial intelligence (AI) tool that can detect carpal tunnel syndrome (CTS), a nerve-related disorder. CTS occurs when the median nerve in the wrist is compressed, leading to symptoms such as numbness, tingling, or pain. The AI tool uses ultrasound videos to identify the median nerve and accurately diagnose CTS.
CTS is a common nerve-related disorder that affects individuals who perform repetitive hand movements, such as office staff, assembly line workers, and sportspersons. Currently, doctors use ultrasound to visualize the median nerve and assess its size, shape, and any abnormalities. However, interpreting ultrasound images and videos can be challenging, especially in regions where the boundaries of the nerve are not clear.
To address this challenge, the researchers at IISc collaborated with Aster-CMI Hospital to develop an AI tool based on a machine learning model with a transformer architecture. The model was originally designed to detect multiple objects in YouTube videos but was modified to focus on tracking the median nerve in ultrasound videos. The researchers collected and annotated ultrasound videos from both healthy participants and individuals with CTS to train the model.
The AI tool successfully segmented the median nerve in individual frames of the ultrasound video, allowing for accurate diagnosis of CTS. Additionally, the tool was able to measure the cross-sectional area of the nerve, a crucial factor in diagnosing CTS. Traditionally, this measurement is performed manually by a sonographer, but the AI tool automates the process and provides real-time measurements with over 95% accuracy at the wrist region.
The AI tool has the potential to revolutionise the diagnosis and treatment of CTS. By automating the detection and measurement process, it can assist doctors in providing timely and accurate diagnoses. Furthermore, the tool's ability to track the median nerve throughout ultrasound videos can aid in treatments that require local anesthesia or nerve blocks for pain relief.
By leveraging machine learning and ultrasound videos, the tool can accurately detect carpal tunnel syndrome and provide real-time measurements of the median nerve. This technology has the potential to improve the diagnosis and treatment of CTS, benefiting individuals who are affected by this common nerve-related disorder.
