Knowledge Vault 3/77 - G.TEC BCI & Neurotechnology Spring School 2024 - Day 8
Decoding of brain signals for brain-computer interfaces at the NTLab
Javier Mauricio Antelis Ortiz, Tecnologico de Monterrey, Campus Guadalajara (MX)
<Resume Image >

Concept Graph & Resume using Claude 3 Opus | Chat GPT4 | Llama 3:

graph LR classDef main fill:#f9d4d4, font-weight:bold, font-size:14px; classDef bci fill:#d4f9d4, font-weight:bold, font-size:14px; classDef systems fill:#d4d4f9, font-weight:bold, font-size:14px; classDef challenges fill:#f9f9d4, font-weight:bold, font-size:14px; classDef summary fill:#f9d4f9, font-weight:bold, font-size:14px; A[Javier Mauricio] --> B[Antelis' BCI research update. 1] A --> C[Develop BCIs for
mobility, communication. 2] C --> D[Collaborations, funding
from Mexico. 3] A --> E[P300 BCI with
robotic orthosis. 4] E --> F[Visual stimuli, user
counts, BCI detects. 5] E --> G[High accuracy in
healthy, ALS subjects. 6] G --> H[No significant differences
between groups. 7] A --> I[Motor imagery BCI
for SCI, FES. 8] I --> J[3D interface, imagined
movement, FES control. 9] I --> K[Tested on healthy,
SCI subjects. 10] K --> L[Both successful, SCI
needed more time. 11] K --> M[Correlations: ERD, activation
time, performance. 12] A --> N[Motor imagery BCI,
VR game, stroke. 13] N --> O[Imagined movements launch
powers, battle monsters. 14] N --> P[Training, testing phases,
cues, feedback. 15] N --> Q[Tested 15 healthy
subjects, variability. 16] Q --> R[Engaging, motivating improvement
over non-VR. 17] A --> S[Challenges: time, equipment,
hospital settings. 18] I --> T[Sham-controlled design
validates BCI effects. 19] A --> U[Hybrid BCIs: P300,
SSVEP, motor imagery. 20] E --> V[Early-stage ALS, artifact
rejection for muscles. 21] A --> W[Prosthetic hand application
from hackathon. 22] E --> X[P300 chosen for stability
over motor imagery. 23] A --> Y[Standard electrodes, no
location comparisons. 24] A --> Z[Several BCIs developed,
promising patient results. 25] class A main; class C,D bci; class E,F,G,H,I,J,K,L,M,N,O,P,Q,R,T,U,V,W,X,Y systems; class S challenges; class Z summary;


1.- Javier Antelis from Tecnologico de Monterrey in Guadalajara, Mexico presents an update on his lab's research on brain-computer interfaces (BCIs).

2.- The lab aims to develop BCIs to help people with reduced mobility and communication, using mainly EEG signals.

3.- Projects involve collaborations with hospitals and other institutions in Mexico and Colombia, with funding from Mexico's National Science Council.

4.- One system is a P300-based BCI integrated with a robotic hand orthosis, tested on healthy subjects and ALS patients.

5.- The P300 system uses visual stimuli that the user counts to make selections, which are detected by the BCI and control the orthosis.

6.- In experiments, both healthy subjects and ALS patients achieved high accuracy in controlling the orthosis with the P300 BCI.

7.- No significant differences were found in performance metrics between the healthy and ALS groups using the P300 orthosis BCI system.

8.- Another system is a motor imagery BCI for spinal cord injury patients to control functional electrical stimulation (FES) for hand grasp.

9.- The motor imagery FES BCI uses a 3D visual interface the user imagines moving to control FES opening/closing of their paralyzed hand.

10.- The motor imagery FES system was tested on healthy subjects and 4 spinal cord injury patients as part of a neurorehabilitation study.

11.- Both healthy and SCI subjects could successfully control the FES with motor imagery, but SCI patients needed more time for activation.

12.- Correlations were found between neural markers of motor imagery (ERD) and FES activation time and online BCI performance.

13.- A third system presented is a motor imagery BCI controlling a virtual reality game, aimed at engaging neurorehabilitation for stroke.

14.- In the VR motor imagery BCI game, imagined left/right hand movements launch fire/ice powers to battle monsters in an immersive environment.

15.- The VR BCI game has separate training and testing phases, with cues for left/right imagery and visual feedback in the VR headset.

16.- So far the VR motor imagery BCI game has been tested with 15 healthy subjects, with some achieving high accuracy but variability.

17.- Surveys showed the healthy subjects found the VR game engaging and motivating, an improvement over previous non-VR BCI experiments.

18.- Challenges in bringing BCIs to patients in hospitals include time constraints, equipment issues, and differences from controlled lab settings.

19.- For the SCI FES study, a sham-controlled design was used with random FES activation in one group to validate BCI-contingent effects.

20.- Some work is being done on hybrid BCIs combining P300, SSVEP and motor imagery, but so far only with healthy subjects.

21.- The ALS patients tested were in earlier disease stages with some residual movement; artifact rejection dealt with unintended muscle activity.

22.- A prosthetic hand project was a favorite application from a recent BCI hackathon hosted by the lab.

23.- P300 was chosen over motor imagery for the ALS orthosis system because it is a more stable and established BCI approach.

24.- Standard electrode locations have been used; no comparisons of alternative locations for P300 vs motor imagery have been done.

25.- In summary, the lab has developed several BCI systems for communication and rehabilitation applications, with promising results in initial patient tests.

Knowledge Vault built byDavid Vivancos 2024