Reconstruction of ECoG signals in response to visual stimuli to decipher the function of brain regions involved in visual processing using a model based on convolutional and regression networks.

The visual system is one of the most sophisticated complex systems in our body, and it plays a crucial role in enabling us to perceive the world around us. When we see images, we send visual information from the eyes to different parts of the brain and various routes transmit visual information and processing. The purpose of this study is to ascertain whether it is possible to reconstruct brain signals directly from visual stimuli using deep neural networks. In order to simulate the visual routes in the brain, we implemented deep neural networks (DNNs) with the objective of predicting the electrocortical data of the whole brain of the Subjects. In this study, we employed an advanced methodology that utilized convolutional neural networks to decode the electrical activity of the brain during the processing of visual data. A convolutional neural network is employed to extract relevant features from the image, which are then fed to a deep regressor for the prediction of the electrocortical data of the subject in that trial. The results demonstrated that brain signals could be reconstructed directly from visual stimuli signals on the trial with acceptable efficiency. Furthermore, neural routes in the brain could be simulated via DNNs. This model could facilitate a deeper understanding of human vision and enhance our comprehension of data processing within the brain.

Keywords:

Convolutional Neural Networks (CNN) , Electrocorticography (ECoG) , Regressor , Vision