Pablo Varona

This paper introduces ExSEnt, a novel entropy-based measure for dementia detection using EEG, showing it achieves comparable classification performance to deep learning methods with a simpler logistic regression model.

The paper examines neuromuscular dynamics during a finger movement task, finding that fast responders show distinct EEG and EMG patterns linked to motor inhibition and activation.

The paper shows how dynamic synapses and intrinsic subthreshold oscillations together enable neurons to recognize specific temporal patterns in spike trains, with robustness to noise and synaptic strength variations.

This paper develops a computational model of the electromotor command system in electric fish that reproduces pulse interval sequences across synthetic and real data.

This paper develops algorithms for automatically adapting model neurons and connections to build hybrid circuits with living neural networks, enabling closed-loop communication and system dynamics characterization.

The paper identifies dynamical invariants in neural activity sequences that maintain robustness and flexibility in central pattern generator circuits.

RTHybrid provides a standardized, open-source real-time software library for building hybrid circuits with living neurons, comparing real-time operating systems for performance and usability.