EEGtofMRI: Learning to translate from fMRI to EEG
Unibrowser: Assistive UI Navigation with Brain-Computer Interfaces (Python/MATLAB)” excerpt: “This is an early concept personal project to learn a generalised translation from EEG to parcillated cortical fMRI using simultaneous EEG/fMRI recordings and comparing different ANN architectures. According to the works of Grooms et al. and others, EEG correlates weakly with cortical fMRI BOLD signal in particular, low frequency and alpha bands. The project is still in its early stages and more sophisticated models (e.g. Deep LSTM networks) and features (e.g. EMD-derived low frequency IMFs etc.) are needed to improve it. Improvement in these models could give medical professionals a real time view of deeper neural activity during surgery and help us to uncover the link between neuronal processing and oxygenation. 