COMPUTATIONAL NEUROSCIENCE
We know just a little about our brains. In this research area we try to approach brain-related information considering both traditional neural data, such as electroencephalographic signals, and data that allow a better understanding on brain conditions, such as speech.
We are also very attentive to treating novel technologies not only from the perspective of computer scientists and as tools ameliorating people's lives, but also as potential threats to fundamental human rights.
TMS-EEG and brain data analysis
To learn more contact
We mainly work on electroencephalographic (EEG) data to access neural information with a rigorous processing pipeline and better understand the connectivity between brain areas.
We are collaborating with Leonor Josefina Romero Lauro, Associate Professor at the Department of Psychology @unimib, Sofia Fazio, Dr. in Physics, Patrizia Ribino and Maria Mannone, Researchers at ICAR-CNR.
Particularly, we are focusing on the following applications ...
Concerning transcranial magnetic stimulation with electroencephalogram (TMS-EEG), we are mainly working to propose novel
- pre-processing pipelines;
- learning models to predict if patients are respondent or not to a specific stimulation;
- how to model neural connectivity.
We are currently proposing physics informed operators to understand neural correlations and brain connectivity in case of episodic events (e.g., epileptic spikes) and neurodegenerative diseases.
Speech analysis of patients with psychotic disorders
To learn more contact
The characteristics of the voice and vocabulary used by people with mental disorders can be important clues for identifying and predicting the onset or worsening of pathology. The study of the speech patterns can therefore be used to facilitate early intervention through telemedicine or to provide help to physicians as decision support or patient monitoring.
This work is being followed in the framework of the DIPPS project.
Particularly, we are focusing on the following applications ...
Investigating speech patterns that can allow to assess symptom severity in people with psychotic disorders.
Definition of machine learning or deep learning models able to automatically recognize and predict the mental disorder by acoustic and linguistic features of speech.