.jpg)
Current research
My research investigates the neurophysiological basis of auditory predictive coding in humans, in the context of the AudioPred Project. I aim to understand how expectancy and surprise shape perception, and what neuronal coding and dynamical principles underlie this.
One line of this work uses intracranial EEG local field potentials and single-unit recordings during tasks with sound streams that alternate between expected and unexpected stimuli with different deviating features. This lets us ask whether specific neurons code for expectancy and violated expectations, and what role rhythms play in carrying the different streams of information that underlie perceptual inference.
A second line uses stereo-EEG (SEEG) high-frequency activity (HFA) to study responses to omitted sounds.Omissions evoke purely internally generated responses, which lets us track prediction and prediction error signals across the auditory hierarchy without any sensory input.
I am also involved in projects on cognitive maps: how latent representations of abstract auditory spaces are encoded across the hippocampal formation, at the single-neuron and population level.
Academic interests
- Predictive coding
- Cognitive maps
- Brain Rhythms
- Computational Neuroscience
- Epilepsy
- Nonlinear dynamics
Background
I have a Bachelor's degree in Control Engineering, and MSc and PhD degrees in Electrical Engineering from the Federal University of Minas Gerais (UFMG) - Belo Horizonte, Brazil. My previous works involved the use of probing stimuli to investigate abnormal neuronal activity in epilepsy. This included the use of steady-state auditory responses in a rat strain susceptible to seizures (Wistar audiogenic rat), and later with epilepsy patients with depth electrode implants for presurgical clinical evaluation. Using a seizure-generating computational model, I also aimed to evaluate the use of stimulation for revealing early-warning signs that may help predict the occurrence of epileptic seizures. In other projects, I use delay differential analysis (DDA), a framework based on delay embedding theory, for speech decoding and investigating seizure dynamics.
Awards
- 2020 Society for Neuroscience Latin American Training Program fellow
Courses
- 2024 Neuromatch Academy NeuroAI course
- 2021 Neuromatch Academy Computational Neuroscience - Project Teaching Assistant
- 2019 Visiting scholar program by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES-BR) at Massachusetts General Hospital
- 2018 VII Latin American School on Computational Neuroscience (LASCON), São Paulo, Brazil
Collaborators
- Prof. Eduardo Mazoni Andrade Marçal Mendes - Universidade Federal de Minas Gerais (UFMG), Brazil
- Prof. Márcio Flávio Dutra Moraes - Universidade Federal de Minas Gerais (UFMG), Brazil
- Dr. Flávio Afonso Gonçalves Mourão - Universidade Federal de Minas Gerais (UFMG), Brazil
- Dr. Claudia Lainscsek - Salk Institute for Biological Studies, La Jolla, CA
- Dr. Lindy Comstock - University of California, Los Angeles
- Dr. Sydney S Cash - Massachusetts General Hospital, Boston, MA