Giuseppe Gallitto

This paper introduces RLBF, a framework for real-time fMRI-based adaptive stimulus generation using reinforcement learning, validated in a study optimizing visual stimuli to enhance V1 responses.

This paper evaluates the replicability of diffusion-weighted MRI-based brain-behavior models and finds moderate replicability with small sample sizes, but higher replicability for trait-like phenotypes.

The paper proposes an adaptive splitting method for external validation of machine learning models to improve generalizability and replicability in predictive modeling studies.

This study shows that statistical learning of non-relevant audio-visual stimuli induces predictive neural responses and error signals in the brain.

This paper introduces fcANNs, a model of brain dynamics based on functional connectivity, showing it can reconstruct and predict brain activity in rest, tasks, and disease states.