electrocorticography -based brain computer interface – the seattle experience
DESCRIPTION
Electrocorticography -Based Brain Computer Interface – The Seattle Experience. Rough Comparison. Task. Control vertical position of cursor to hit the target. Horizontal Speed = 1 screen width / 5.5 seconds. Task Interface. Decoder. Output (25 Hz): “up” or “down” magnitude. - PowerPoint PPT PresentationTRANSCRIPT
![Page 1: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/1.jpg)
Electrocorticography-Based Brain Computer Interface – The Seattle Experience
![Page 2: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/2.jpg)
Rough Comparison
EEG ECoG Implanted Arrays
Invasiveness Low Medium HighEMG Noise High Medium Low
Risk Low Medium HighStability High Medium Low
Spatial Resolution 1 cm 0.01 cm 0.001 cm
Price $100s gazillion? buzillion?
![Page 3: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/3.jpg)
Task
Control vertical position of cursor to hit the target
Horizontal Speed= 1 screen width/ 5.5 seconds
![Page 4: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/4.jpg)
Task Interface
DecoderInput (1,000 Hz):- 64 ECoG electrodes
Human User
Output (25 Hz):- “up” or “down”- magnitude
Visual feedback
![Page 5: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/5.jpg)
System Overview
• ECoG electrode placement• Decoding• Learning (Model)• Experiment
![Page 6: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/6.jpg)
ECoG Placement
![Page 7: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/7.jpg)
Decoding
• For each user U and user action A– Feature functions f(x)– Feature weights w
• Output is linear combination of feature functions
• How to choose features?• How to weight features?
![Page 8: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/8.jpg)
Feature Selection
Rest state
Action state
![Page 9: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/9.jpg)
Feature Selection + Learning
• Training data = <signal, action state> pairs– Signal = input from electrodes– Action state = “performing action” or “not”
• Possible features = amplitude of {electrode1, electrode2, …} x {freq1, freq2, …}
• Rank features using autoregressive model• Choose top K• Weights from autoregressive model (?)
![Page 10: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/10.jpg)
Experiment
1. Offline training to learn features, weights2. Online development testing3. Online feature, weight adjustment4. Final round of testing
![Page 11: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/11.jpg)
Interesting Observation
• Offline (no feedback) looks different than online (with feedback)
![Page 12: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/12.jpg)
Results
![Page 13: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/13.jpg)
Results (from related paper)
![Page 14: Electrocorticography -Based Brain Computer Interface – The Seattle Experience](https://reader035.vdocuments.us/reader035/viewer/2022070422/56816388550346895dd475b2/html5/thumbnails/14.jpg)
Conclusions
• Users can control a 1d cursor with ECoG • Closed loop looks different than “open loop”• Experimenting with epilepsy patients is hard