fNIRS-based Neurorobotic Interface for gait rehabilitation.
Rayyan Azam Khan, Noman Naseer, Nauman Khalid Qureshi and 3 others
PMID 29402310WHAT IT FOUND
Nine healthy men produced brain signals recorded with fNIRS while walking on a treadmill.
A classifier reached 86.7% accuracy in the signal-decoding tests, and a simulated prosthetic leg reduced position error in under 2.5 seconds. No patients were treated.
Key findings
01SVM with hrf filtering produced accuracies of 77.5, 72.5, 68.3, 74.2, 73.3, 80.8, 65, 76.7, and 86.7% across the nine subjects.
02For offline BCI, SVM showed greater statistical significance than the other classifiers, with p < 0.01.
03In simulation, the controller minimized position error in less than 2.5 s.
STILL TO COME
How it was doneWhat they found
Read the rest of this summary
You get three full summaries a month, free, and we do not ask for a card. Search, the TL;DRs and your library stay unlimited either way.
What it does not show
Only nine healthy men were studied, so the results do not show what happens in people with stroke, spinal cord injury, or amputation. No clinical gait outcomes, safety outcomes, or patient rehabilitation measures were reported. The prosthetic leg was simulated, not physically tested with a patient. The fNIRS signal measures blood flow changes rather than nerve activity directly, and the paper states this creates delay. During rest intention, the leg held its last position rather than the initial position. The paper does not describe randomisation, blinding, or dropout handling. Classification accuracy varied between subjects, and the authors note head shape and scalp-cortex distance may affect signals.
The easy way to misread this
Do not read this as evidence that a brain-controlled prosthetic leg improves walking in patients. The study used nine healthy men and a simulated leg, and it did not measure clinical gait outcomes.