Design of MRI structured spiking neural networks and learning algorithms for personalized modelling, analysis, and prediction of EEG signals

Abstract This paper proposes a novel method and algorithms for the design of MRI structured personalized 3D spiking neural network models (MRI-SNN) better analysis, modeling, prediction EEG signals. It gradient-descent learning algorithm integrated with spike-time-dependent-plasticity algorithm. The capture informative personal patterns interaction between channels, contrary to single signal modeling methods or spike-based approaches which do not use data pre-structure model. proposed can only learn model accurately measured data, but they also predict signals at locations that correspond non-monitored brain areas, e.g. other from where has been collected. is first study in this respect. As an illustration method, MRI-SNN are created tested on two subjects. result accuracy understanding than traditional due information integration. interpretable facilitate related processes. approach be applied across studies such as epilepsy, peri-perceptual activities, brain-computer interfaces, others.