My PhD research consists of the processing of signals from a 14-electrode EEG system, connected to immersive glasses that allow for a realistic visual experience and for the investigation of the brain network in order to identify signal features corresponding to different perceptive and cognitive stimuli. The aim of the research is to implement a procedure that identifies correspondences among EEG signals and chaotic attractors. The chaotic attractors can be defined as a trajectory of a dynamical system, contained in a defined volume of phase space. A dynamical system can have chaotic behavior, i.e. an organized (but not periodic) behavior sensitive to the initial conditions. EEG signals can be considered dynamical systems. In this work a custom Artificial Neural Network (ITSOM) processes individual signals or many signals simultaneously. The sequence of the ITSOM winning nodes tends to repeat itself creating a time series of chaotic attractors. The ITSOM attributes similar codes to attractors emerging from similar brain states, perceptions and emotions. These attractors are isomorphic to the attractors in which the corresponding dynamical system (the signal time series) is evolving and univocally characterize the input element that produces them. If the attractors are chaotic, this means that the signals are individually self-organized or, by examining more signals together, there is a form of coherence among signals. The ITSOM network memorizes the time series of the winning nodes. The cumulative scores for each input are normalized following the z standardized variable distribution. Attractors are labeled with a binary code that univocally identifies them, and the flexibility of the Artificial Neural Network allows attributing the same codes to similar dynamical events. During the experiment, the subject is looking at the screen while different shades of colors, yellow, red and blue are displayed. Each stimulation lasts five seconds, between stimuli there is a black screen, used to reset the previous color stimuli. The collected results show, as forecast, many correspondences among binary codes coming from similar stimuli. The thesis provides a detailed description of these results.
ARTIFICIAL INTELLIGENCE APPLIED TO THE STUDY OF CONSCIOUS PERCEPTIVE STATES / M. Musumeci ; RITA MARIA ROSA PIZZI. - Milano : Università degli studi di Milano. DIPARTIMENTO DI INFORMATICA Giovanni Degli Antoni, 2019 Feb 01. ((31. ciclo, Anno Accademico 2018.
|Titolo:||ARTIFICIAL INTELLIGENCE APPLIED TO THE STUDY OF CONSCIOUS PERCEPTIVE STATES|
|Data di pubblicazione:||1-feb-2019|
|Parole Chiave:||ARTIFICIAL INTELLIGENCE; CONSCIOUSNESS; BRAIN; MIND; NEURAL NETWORKS|
|Settore Scientifico Disciplinare:||Settore INF/01 - Informatica|
Settore M-PSI/02 - Psicobiologia e Psicologia Fisiologica
Settore BIO/09 - Fisiologia
|Citazione:||ARTIFICIAL INTELLIGENCE APPLIED TO THE STUDY OF CONSCIOUS PERCEPTIVE STATES / M. Musumeci ; RITA MARIA ROSA PIZZI. - Milano : Università degli studi di Milano. DIPARTIMENTO DI INFORMATICA Giovanni Degli Antoni, 2019 Feb 01. ((31. ciclo, Anno Accademico 2018.|
|Digital Object Identifier (DOI):||http://dx.doi.org/10.13130/musumeci-marialessia_phd2019-02-01|
|Appare nelle tipologie:||Tesi di dottorato|