Skip to main content

Paper out: AI method for brain graphs and psychosis detection

The workflow for the GCN model used in the paper.  Brain scans are converted into functional and structural connectivity networks. Then a GCN models is applied to classify the networks as patient or control.

We have a new article out on the classification of psychosis spectrum disorders using graph convolutional networks. A graph convolutional network (GCN) is a type of artificial intelligence (AI) that uses networks (also called graphs) to as inputs to do common AI tasks like classification and feature mapping to detect important brain regions.  

The workflow for the GCN model used in the paper.  Brain scans are converted into functional and structural connectivity networks. Then a GCN models is applied to classify the networks as patient or control.