How few channels suffice? A hierarchical validation framework for minimal-channel wearable sEMG in upper-limb rehabilitation gesture recognition.

All 24 participants were healthy young adults, average age 21.3. Nothing here was tested on people with stroke, nerve injury or any other neuromotor impairment, and the authors say the layout may need to change in those groups. Only eight discrete gestures were tested, covering wrist and hand movement but not the full range of daily activities, and continuous proportional control such as holding a steady grip was not tested at all. Recordings were analysed offline with a fixed Ag/AgCl electrode layout. Real-time closed-loop control, dry or textile electrodes and other mounting positions were not assessed, so latency, battery drain and electrode shift during daily use are unmeasured. The electrode pool had only four forearm sites and two upper-arm sites, so this study cannot say whether eight electrodes packed onto the forearm would beat four. The whole result depends on very clean signals. When the researchers deliberately degraded the signal, the small electrode sets lost accuracy far faster than the six-electrode array, so the four-electrode figure may not hold on cheaper hardware.

The work was funded by the Young Scientists Fund of the National Natural Science Foundation of China, the Taishan Scholars Project (Young Expert Program), the Shandong Excellent Young Scientists Fund Program (Overseas), and the Open Project Fund of the International Joint Research Center for Perception and Control of Intelligent Rehabilitation Systems of Sichuan Province. The text supplied does not include a competing-interests declaration.

Do not read this as a device your patients can use. The 94.44% figure comes from 24 healthy young adults performing eight set gestures in a lab and analysed offline. No stroke or nerve-injury patient was tested, and the authors say the four-electrode layout still needs validating in those groups before clinical deployment.