EEGG: An analytic brain-computer interface algorithm
preprintposted on 07.01.2021, 04:01 by Gang Liu, Jing Wang
Objective. A black box called brain-computer interface (BCI) model is used to identify another black box, the brain. However, one black box cannot explain another black box. This paper presents the first analytic "white box" brain-computer interface algorithm named EEGG.
Approach. Independent and interactive effects of neurons or brain regions can fully describe the brain. This paper constructed a relationship model that extracted the independent and interactive features of EEG for intention recognition and analysis using ResDD, a novel dendrite module of Gang neuron. A total of 4,906 EEG data about motor imagery (MI) of left-hand movements and right-hand movements from 26 subjects were obtained from GigaDB. Firstly, we explored EEGG's generalization ability according to cross-subject accuracy. Secondly, we transformed the EEGG model into a relationship spectrum expressing independent and interactive effects of brain regions. Then, the relationship spectrum was verified through the known ERD/ERS phenomenon. Finally, we explored the previously unreachable further analysis based on a BCI model.
Main results. (1) EEGG was more robust than typical "CSP+" algorithms for the poor quality EEG data [AUC:0.825±0.074(EEGG)>0.745±0.094(CSP+LDA)/0.591±0.104(CSP+Bayes)/0.750±0.091(CSP+SVM), p<0.001]. (2) The transformed EEGG model showed the known ERD/ERS phenomenon. (3) Interestingly, EEGG showed that the interactive effects of brain regions put a brake on ERD/ERS effects for classification (p<0.001). This means that generating fine hand intention needs more centralized activation of the brain.
Significance. EEGG implies that, henceforth, not only can BCI be used for recognition but also analysis.