Sina Mansour L.

97% of 184 high-impact biomedical AI studies compare cross-validated models with tests assuming fold independence; in 69% of re-analyzable studies a significant comparison does not survive correction.

This paper presents a simulation framework to evaluate EEG source localisation methods by generating biologically plausible seizure dynamics and EEG signals with known ground truth.

This paper uses structural eigenmodes of the brain to improve EEG source localization in epilepsy, finding geometric eigenmodes slightly outperform connectome eigenmodes.

The paper introduces a deep learning model to generate individual connectomes from personal attributes and shows it can accurately replicate empirical connectomes from the UK Biobank.

This study evaluates the accuracy of reconstructing cortical functional maps using subject-specific geometric and connectome eigenmodes, finding that spatially smooth eigenmodes best explain functional data.

This paper introduces DELSSOME, a deep learning framework that accelerates biophysical brain circuit model optimization, enabling faster parameter estimation and lifespan trajectory analysis of cortical E/I ratios.

This study identifies and addresses spurious correlations in surface-based fMRI data caused by uneven vertex spacing in processing pipelines.

This study maps thalamocortical connectivity in infants using diffusion MRI and finds a conserved anterior/posterior gradient that matures during gestation.

The paper re-evaluates the comparative explanatory power of geometric and connectome eigenmodes in predicting brain activity, finding that connectome eigenmodes can achieve comparable accuracy when improved connectome mapping is used.

The paper provides a large-scale connectome resource for 40,000 UK Biobank participants, including structural and functional brain networks with multiple parcellations and preprocessing techniques.

The paper uses intracranial EEG data to show that structural connectivity models can predict the propagation of electrical brain stimulation across the cortex.

The paper shows that adding models of indirect neural communication to structural brain networks better aligns their modular organization with functional systems.

This paper introduces Connectome Spatial Smoothing (CSS) to improve the reliability and sensitivity of structural connectomes by applying spatial smoothing to tractography data.

The paper validates that rsfMRI temporal complexity is reproducible and correlates with fluid intelligence across multiple brain networks.

The paper examines how brain age gaps relate to psychopathology dimensions in youth, finding significant associations with psychosis, obsessive-compulsive, and general psychopathology.