Title: Cognitive Control Signals for Neural Prosthetics
1Cognitive Control Signals for Neural Prosthetics
- Sam Musallam
- The Andersen Lab
2Cognitive-based rather thanmotor-based
- May require fewer cells, less invasive
- Cognitive variables-expected value
- Hierarchical control
PMC
3We are currently implanting 160 electrodes in
PRR and in DPMC
- 32 electrode arrays - Made from platinum and
iridium - 80 microns at the shaft - 2-3 micron
tip
4Arrays are Implanted in the Medial Intraparietal
Sulcus and Dorsal Premotor Cortex
5Arrays are Implanted in Medial Intraparietal
Sulcus
6Decoding Goals E
Reach Trials
Decode Trials
7Reach Trials
8Neuron example from parietal cortex
9Decoding Goals E
Reach Trials
Decode Trials Bayes rule
10Brain Control Trials
11Example of feedback performanceParietal
Fixation maintained throughout trial.
12Brain control trials using neurons from DPMC
75
13Frozen or adaptive database? Doesnt seem to
matter
14At 100 ms, predictions are still significantly
above chance
15Overall performance improves with time
Mutual Information
Monkey Performance
16Identical but different trials can elicit
different firing rates
17In a single session, feedback performance
fluctuates
6 Targets
Chance
18Type of reward changes the tuning of neurons in
PRR
Blackorange juice Red water
19Tuning is also enhanced for increased probability
and magnitude of reward
20Our decoding ability improves with preferred
reward condition
21MI increases with preferred reward trials
22Overall decode improvement
23Decoding the value of reward
24Decoding Direction and Reward
25More neurons improve the decode
26This behavior is advantageous for prosthetic
control
- Optimize control of prosthetic devices under
given decoding constraints - counter neuronal sample biases(surgical
placement of electrodes, etc.) - counter signal non-stationarities(changes in
tissue, electrode characteristics, etc.) - allow multiple tool use(update or variable
functionality of prosthesis)
27What about trajectories?
2880 accuracy to 8 targets
Mean time taken to reach target 700
ms. Fixation point is randomized.
29Conclusion
- Cognitive neural activity not directly related to
visual input or motor output can be used for as
prosthetic control signals. - Cognitive signals can also give us information
about the patients preference or mood - Signals in higher level areas like the PPC and
DPMC can be used to guide trajectories as well
30- Brian Corneil
- Hans Scherberger
- Bradley Greger
- Grant Mulliken
31Thank You
32(No Transcript)