Roberto Carlos Sotero

This paper introduces a statistical meta-aggregation method for EEG classification in Parkinson's disease, improving subject-level classification performance by aggregating epoch-wise predictions more effectively.

This paper introduces a metabolic cost asymmetry as a structural regularizer for recurrent spiking neural networks, showing it improves performance and reduces energy use.

This paper introduces a physics-informed neural network to recover subject-specific travelling-wave structure in resting-state fMRI by incorporating anatomical priors.

This paper uses Stochastic Actor-Oriented Models to analyze brain network changes with aging, identifying factors driving connectivity dynamics in functional and structural networks.

This paper estimates the excitatory-inhibitory balance from ECoG data using Physics-Informed Neural Networks, finding reduced long-range and excitatory short-range connections under anesthesia.

This paper uses Physics-Informed Neural Networks to estimate directed functional connectivity from rsfMRI data, revealing sex-specific differences in autism and neurotypical brains.

This paper estimates parameters in brain dynamics models from resting-state fMRI data using Physics-Informed Neural Networks, revealing differences in brain network structure between neurotypical individuals and those with ASD.

This study improves EEG source localization by using high spatial-frequency components from fMRI, showing better performance than traditional methods.

This study introduces an adaptive signed random walk model to analyze brain networks with both positive and negative connections, showing it captures more interactions than classical methods.

This paper improves ASD classification using random walks-informed feature extraction and selection, achieving 76.54% accuracy with SVM.

This paper introduces a noise removal method for EEG gamma-band signals using Holo-Hilbert spectral analysis to reduce artifacts while preserving the original signal.

This paper introduces a method to detect brain network communities at different temporal scales by analyzing information flow through oscillatory modes and clustering nodes based on their co-occurrence.

This study examines how including anti-correlations in graph theory features affects SVM classification performance for autism spectrum disorder using resting-state fMRI data.

The paper introduces a new measure of brain network complexity using random walks and finds that the positive functional connectivity network has higher complexity than other network types.

This paper applies spatiotemporal empirical mode decomposition to resting-state fMRI to remove global signal noise, resulting in more accurate functional connectivity matrices.