Design and Implementation of EEG-fNIRS Based Brain-Computer Interface Paradigm for Musical Imagery

In this experiment, EEG-fNIRS simultaneous acquisition technology was used to design and implement a brain-computer interface paradigm for musical imagery, and the self-designed experimental paradigm was utilized for sampling with a size of 40 college student subjects, data processing, time-frequency domain feature extraction, statistical analysis based on EEG, fNIRS in the resting, white noise, listening, and recall phases, and neurovascular coupling were carried out. Finally, music recognition based on six songs was performed by machine learning, resulting in the best recognized song as well as its features. In the future, the construction of a multimodal database based on simple elements of musical imagery is planned.