Classification of upper limb center-out reaching tasks by means of EEG-based continuous decoding techniques.
Andrés Úbeda, José M Azorín, Ricardo Chavarriaga and 1 others
PMID 28143603WHAT IT FOUND
EEG signals from healthy adults doing reaching tasks could decode movement direction, but accuracy was not high enough for real-time cursor control.
Simplifying the task to classify which target was reached gave higher accuracy, especially when fewer directions were compared.
Key findings
01Active reaching was decoded from EEG significantly above chance for position and velocity.
02Classifying which target was reached yielded high performance and was significantly above chance across configurations.
03Passive reaching was not decoded above chance, while active reaching was significantly better than passive reaching.
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
The study tested healthy adults, not patients with stroke or spinal cord injury. The authors say caution is needed because the sample size was small and larger datasets are required. The decoding was tested by splitting each person's own recordings into training and test parts, not as a real-time system a patient could use. The paper reports movement decoding and classification accuracy, not recovery of arm function or daily activities.
Declared interests
The work was funded by Ministerio de Economía y Competitividad, Conselleria d'Educació, Cultura i Esport of Generalitat Valenciana, and Schweizerischer Nationalfonds zur Förderung der Wissenschaftlichen Forschung. The paper lists funders but does not state competing interests.
The easy way to misread this
Do not read this as evidence that EEG-based brain-machine interfaces improve rehabilitation. The work was done in healthy adults, measured decoding accuracy rather than patient function, and the authors say the continuous decoding was not accurate enough for real-time cursor control.