PMID-17978021[0] Biomimetic Brain Machine Interfaces for the Control of Movement.
- images/482_1.pdf
- describe structured models that include arm information & 'plant' dynamics.
- current methods ignore the dynamics of the musculoskeletal system. Want to mimic natural arm movement.
- To this end used a kinarm with a paralyzed monkey.
- obtained real-time prediction of joint force, torque, and EMG
- Concerning quality of prediction: they use fraction of movement variance that can be accounted for (FVAF) which, though google does not seem to know much about it, is probably the same as R^2. but it does not look that great:
- 0.61 - 0.65 for torque prediction
- 0.70 - 0.75 for EMG prediction once again, the limitation is the recording technology.
- tested coupling predictions to the freehand FES system - see this crazy news brief
- want to incorporate somatosensory feedback into the BMI.
- they reference a paper from 2008 - huh? The document claims to be written/published in 2007.
____References____
[0] Fagg AH, Hatsopoulos NG, de Lafuente V, Moxon KA, Nemati S, Rebesco JM, Romo R, Solla SA, Reimer J, Tkach D, Pohlmeyer EA, Miller LE, Biomimetic brain machine interfaces for the control of movement.J Neurosci 27:44, 11842-6 (2007 Oct 31) |
|