MovementIntentionDecodingAlgorithms

Neural prosthetics monitor signals from neurons and use these signals to control effectors (cursors on a computer screen, robotic limbs). In order to figure out where to move the effector, one has to decode the neural signals in realtime. Which algorithms have been proposed for this?

Comparisons and review articles

• Kim et al '06. Covers some linear models (Wiener filter, LMS? adaptive filters, gamma filter, subspace Wiener filters) and some nonlinear models (time-delay neural network and local linear switching models).

List of algorithms

• Linear approaches
  • Linear regression
    • the Nagle paper (20 bins of 50ms each) (todo: add ref).
    • The Carmena et al paper where a monkey controlled a robotic arm also used linear regression (100ms bins, 10 taps).
  • (Michael Black’s talk). Haven’t looked at it yet myself, but see talk by Michael Black which “summarize[s] the work at Brown University on real-time neural control of cursor motion”.
• “Black box” nonlinear approaches
  • Support Vector Machine (SVM?)
    • Olson et al '05. Operates on binned spike rasters w/ 100ms per bin.
• Uncategorized
  • • Multivariate Gaussian or Poisson: http://www.nature.com/nature/journal/v442/n7099/full/nature04968.html
  • Switching Kalman filter
    • Wu et al '04. Operates on spike rates averaged over 50ms.
  • Also mentioned in that Black’s presentation:
    • Population Vectors: Georgopoulos et al. (1986), Moran & Schwartz (1999), Taylor et al. (2002)
    • Linear Regression Methods: Wessberg et al.(2000), Serruya et al. (2002), Carmena et al. (2003)
    • Artificial Neural Networks: Wessberg et al.(2000)
    • Bayesian Inference (e.g. particle filter): Gao et al. (2002), Brockwell et al. (2004), Wu et al. (submitted, EMBS)

Refs

• Jose M. Carmena, Mikhail A. Lebedev, Roy E. Crist, Joseph E. O’Doherty, David M. Santucci, Dragen F. Dimitrov, Parag G. Patil, Craig S. Henriquez, and Miguel A. L. Nicolelis. Learning to control a brain-machine interface for reaching and grasping by primates. PLoS Biology, 1(2):193-208, November 2003.
• S-P Kim, J C Sanchez, Y N Rao, D Erdogmus, J M Carmena, M A Lebedev, M A L Nicolelis and J C Principe. comparison of optimal MIMO linear and nonlinear models for brain–machine interfaces. A comparison of optimal MIMO? linear and nonlinear models for brain–machine interfaces</a>. Journal Neural Engineering 3 145-161 (2006).
• Olson, B.P.; Si, J.; Jing Hu; Jiping He. Closed-loop cortical control of direction using support vector machines. IEEE? Transactions on Neural Systems and Rehabilitation Engineering. Volume 13, Issue 1, March 2005 Page(s):72 - 80.
• Wei Wu; Black, M.J.; Mumford, D.; Yun Gao; Bienenstock, E.; Donoghue, J.P. Modeling and decoding motor cortical activity using a switching Kalman filter. IEEE? Transactions on Biomedical Engineering. Volume 51, Issue 6, June 2004 Page(s):933 - 942.

CategoryIo

CategoryTodo