Multi-Mode Brain Network Calculation Method and Device Associated With Structure and Function, Equipment and Medium
Abstract
The invention discloses a structure and function associated multi-modal brain network calculation method, device and equipment and a medium, which are applied to training a brain disease prediction model, the model comprises an associated perception dual-channel generation module, a disease feature regression module, a topological structure discriminator and a time-space joint discriminator, and in the model training process, the time-space joint discriminator is used for predicting the brain disease. The method comprises the following steps: performing multi-level interactive fusion learning on high-order topological characteristics of brain function magnetic resonance data and magnetic resonance diffusion tensor imaging data to obtain a multi-modal time sequence activity signal of each brain region, and combining directional overall causal inference based on a brain region activity structure equation to obtain a multi-modal time sequence activity signal of each brain region; the directional causal effect relationship among the brain regions is described, and multi-modal effective connection calculation is realized, so that the trained model can perform intelligent auxiliary pathological analysis and focus traceability on the neurodegenerative disease patient by using multi-modal effective connection.
CN115813367A
