In a multi-layered neuromimetic network model (supervised deep learning), to obtain as a result the signal (time serie) of a modality A at a point P of an organ, I would like to enter two types of information: time series of signals corresponding to several channels of a modality B on the one hand, and a representation of the desired location (3D) for the point P. Could you help me by suggestions of bibliographical references or good ideas? Thank you.