Giulio Ruffini

cyto7 is a vertex-level cortical-type atlas of the human cortex that improves upon area-level maps by capturing finer type boundaries and validating its structure with multimodal data.

This paper proposes that cross-frequency coupling (CFC) mechanisms in a laminar neural mass model support prediction error evaluation in predictive coding, with implications for disorders like Alzheimer's disease and the effects of serotonergic psychedelics.

This paper introduces a variational autoencoder for detecting seizure onset phases in SEEG data, achieving high recall and interpretable latent features.

This paper analyzes a laminar neural mass model to show it can generate periodic, quasiperiodic, and chaotic oscillations, including cross-frequency couplings, and examines how Alzheimer's disease affects these dynamics.

This paper develops a laminar neural mass model to simulate Alzheimer's disease progression, showing that PV interneuron dysfunction drives early hyperexcitability and later hypoactivity through electrophysiological biomarkers.

This paper uses a whole-brain model to simulate how serotonergic psychedelics affect brain dynamics in Alzheimer's disease, showing restored oscillatory patterns and correlation with serotonin receptor distribution.

This paper introduces Hierarchical Amplitude Modulation (HAM) as a mechanism for hierarchical information processing in the brain, showing how it produces observed spectral features through log-spaced frequency bands and 1/f power laws.

The paper uses whole-brain modeling to analyze brain dynamics during advanced meditation states, finding increased functional integration and near-critical dynamics in later states associated with minimal phenomenal experience.

This paper develops personalized whole-brain models of seizure propagation using multimodal clinical data to simulate and evaluate individualized therapeutic strategies.

This paper introduces a group optimization framework for tES dose planning using representative head models, showing improved consistency and predictive power compared to individualized approaches.

This paper introduces a biologically grounded model that explains both cooperative and competitive interactions in the brain using laminar-specific cross-frequency coupling, validated on fMRI data from 60 subjects.

The paper explores how symmetry and structure in data influence the dynamics and organization of algorithmic agents, linking concepts from group theory, information theory, and neural networks.

This study examines how classical psychedelics alter brain hierarchy and complexity, finding they reduce hierarchy while increasing neural complexity.

The paper explores how psychedelics affect neural dynamics by flattening the neural landscape, increasing complexity, and promoting adaptability through changes in plasticity mechanisms across different timescales.

The paper evaluates biophysical models to characterize electric fields from intracranial electrodes in DRE patients, finding that electrode geometry and tissue properties influence field distribution.

The paper uses computational modeling to identify key brain regions for personalized neurostimulation strategies to improve cognitive interventions in Alzheimer's disease.

This paper shows complex spatiotemporal brain dynamics emerge from transverse instabilities in large-scale neural models.

This paper introduces a physical neural mass model framework to infer oscillatory generator locations from laminar electrophysiological recordings in macaque monkeys.

The paper quantifies how weak electric fields affect neuron models using spherical harmonics, showing dipole dominance and other terms' roles in non-resting states.

This study uses Ising models and algorithmic complexity to analyze fMRI data and finds LSD increases brain complexity and moves it further from criticality.

The paper develops a framework linking neural mass models with physical models to simulate SEEG recordings more accurately by incorporating current source density and volume conductor effects.

The paper compares an exact neural mass model (NMM2) with a heuristic one (NMM1) in the presence of second-order synaptic dynamics, showing NMM1 fails to capture key dynamical features.

This paper introduces a laminar neural mass model that realistically reproduces SEEG recordings during epileptic seizures and demonstrates its ability to capture key electrophysiological features.

This paper investigates how variations in electric field distribution from tDCS affect cognitive-motor outcomes in older adults and shows that optimized multichannel montages can improve targeting and reduce variability.

The paper uses the 2D Ising model to explore links between criticality and complexity, focusing on phase transitions and their implications for neuroscience and information theory.

The paper extends an exact mean-field theory for neural mass models by incorporating dynamic synaptic currents and external electric fields, showing how network connectivity influences response to weak inputs.

This study models ephaptic fields in the human brain across aging, finding they decrease with age and follow gyrification patterns, with higher involvement in sensorimotor and medial regions.

The paper investigates conscious perception by analyzing brain networks during magnetophosphenes, finding distinct roles for alpha and beta band oscillations in integration and segregation.

This study shows that multifocal tDCS modulates resting-state connectivity in older adults, with effects dependent on electric field distribution and baseline connectivity.

StarNEt, a deep learning model, improves MRI segmentation for transcranial electrical stimulation by providing faster and more accurate white and gray matter masks compared to FreeSurfer.