Raju Surampudi Bapi

The paper introduces graph diffusion wavelets to estimate functional connectivity from structural connectivity, achieving high correlation and revealing power-law diffusion scales and frontal pole dominance.

This paper introduces A-GHN, an attention-based fusion of multiple GraphHeat networks to map structural to functional brain connectivity using resting-state fMRI data from the HCP.

mulEEG is a multi-view self-supervised method for EEG representation learning that outperforms baselines and supervised training on sleep-staging tasks.

This paper presents a computational model of reward prediction error computation in the VTA, proposing that GABA signals encode magnitude and timing separately from the Ventral Striatum and Peduncolopontine.

This study quantifies ASD severity by analyzing dynamic functional connectivity variability and modular organization in resting-state fMRI data across different age groups.

The paper introduces a temporal Multiple Kernel Learning model to predict resting-state functional connectivity by characterizing fMRI connectivity dynamics and shows improved performance over existing methods.

This study investigates functional connectivity differences in ASD versus TD using graph-theoretic analysis across two datasets, finding age-related and disease-modulated changes in network properties.

This paper introduces a Hamiltonian formalism to construct Liapunov functions for recurrent neural networks, enabling analysis of emergent properties and psychophysical applications.

This paper proposes a multiple diffusion kernel model to estimate brain functional connectivity from structural data across subjects, showing it outperforms single kernel approaches.