Normative Modeling

Normative modeling in neuroimaging establishes statistical norms of healthy brain structure and function to identify deviations indicative of disease. Current research focuses on improving the accuracy and interpretability of these models, employing advanced architectures like variational autoencoders (VAEs) and transformers, often incorporating multiple imaging modalities to capture disease heterogeneity. This approach offers a powerful tool for early disease detection, personalized diagnosis, and a deeper understanding of neurological and psychiatric disorders by characterizing individual differences in brain patterns relative to a healthy population.