My research aims to understand why individuals differ in their brain organization, cognitive and socio-affective functioning, and vulnerability to neurological and psychiatric disorders. To address this question, my group develops systems-level approaches that integrate multimodal neuroimaging, behavioral phenotyping, exposomic data, and machine learning.

A central focus of my work is the organization of the human brain and its relationship to behavior. Rather than assuming simple one-to-one mappings between individual brain regions and psychological functions, we study brain–behavior relationships as multidimensional patterns distributed across the brain. The hippocampus has served as an important model system in this research. By combining structural MRI, functional connectivity, and molecular imaging, we have characterized its complex organization across development, adulthood, and aging, as well as alterations associated with dementia.

A second major research focus concerns interindividual variability and the challenges of predicting behavior from neuroimaging data. Brain-based predictive models often achieve only modest accuracy, are sensitive to analytical decisions, and may not generalize across datasets or population groups. My research therefore emphasizes replicability, interpretability, and generalizability as essential criteria for neuroscientific machine learning. In particular, we investigate why models perform differently across individuals and subgroups and develop approaches that better represent heterogeneity in brain–behavior relationships.

More recently, my research has introduced an exposome perspective into systems neuroscience. The exposome encompasses the environmental, lifestyle, social, and biological influences experienced by an individual across the lifespan. These influences do not act independently: they form complex, interacting patterns whose timing, duration, and intensity contribute to shaping the brain and behavior. Using large population cohorts, we investigate how exposome-wide profiles (including education, physical activity, sleep, cardiometabolic health, physical robustness, psychosocial conditions, and environmental factors) relate to individual brain phenotypes. Our work has demonstrated that brain health in aging can be predicted from such exposome-wide patterns and that different dimensions of bodily health are associated with partly distinct neurobiological signatures.

This framework is applied to the heterogeneity observed in neurological and psychiatric disorders, particularly dementia and depression. Individuals sharing the same diagnosis can differ substantially in their brain alterations, symptoms, and disease trajectories. We therefore move beyond conventional case–control comparisons toward dimensional and data-driven representations of vulnerability, resilience, and disease expression. 

By integrating the exposome, brain, and behavior, my research ultimately seeks to advance precision neuroscience and contribute to more personalized and preventive approaches to brain and mental health.