Development of Deep Belief Networks to support the diagnosis of psychiatric disorders

• Grant number: 13/05168-7
• Support Opportunities: Scholarships in Brazil - Doctorate
• Start date: July 01, 2013
• End date: December 01, 2016
• Field of knowledge: Interdisciplinary Subjects
• Principal Investigator: João Ricardo Sato
• Grantee: Walter Hugo Lopez Pinaya
• Host Institution: Centro de Matemática, Computação e Cognição (CMCC). Universidade Federal do ABC (UFABC). Ministério da Educação (Brasil). Santo André , SP, Brazil
• Associated scholarship(s): 15/14993-7 - Development of Deep Belief Networks to support the diagnosis of psychiatric disorders from large-scale public databases, BE.EP.DR

Abstract

Neuroscience has served as basis for a better understanding of phenomena related to brain function. Its development helps with various fields of medicine such as psychiatry, which currently still performs a diagnostic work very subjective. So any objective information linked to the neural substrates of psychiatric disorders are of great value as a complement to the conventional diagnostic methods. Several researches seek biological markers, using neuroimaging techniques and genetics. As these investigative techniques create large volumes of data, machine learning is indicated to evaluate small variations related to psychiatric disorders, but most studies done so far achieved modest results, since they work with relatively small samples. This project will be based on one of the most modern models using machine learning, the Deep Belief Networks, which has generated promising results by getting performances superior to other tools. This model allows the implementation of artificial neural networks with multiple hidden layers, obtaining good results in large training bases. This project will introduce this approach in the analysis of markers to support diagnosis of psychiatric disorders using databases with hundreds of individuals. (AU)

Scientific publications

(5)

(References retrieved automatically from Web of Science and SciELO through information on FAPESP grants and their corresponding numbers as mentioned in the publications by the authors)

VIEIRA, SANDRA; PINAYA, WALTER H. L.; MECHELLI, ANDREA. Using deep learning to investigate the neuroimaging correlates of psychiatric and neurological disorders: Methods and applications. NEUROSCIENCE AND BIOBEHAVIORAL REVIEWS, v. 74, n. A, p. 58-75, MAR 2017. (13/05168-7)

PINAYA, WALTER H. L.; GADELHA, ARY; DOYLE, ORLA M.; NOTO, CRISTIANO; ZUGMAN, ANDRE; CORDEIRO, QUIRINO; JACKOWSKI, ANDREA P.; BRESSAN, RODRIGO A.; SATO, JOAO R.. Using deep belief network modelling to characterize differences in brain morphometry in schizophrenia. SCIENTIFIC REPORTS, v. 6, DEC 12 2016. (13/05168-7, 13/10498-6)

DE MOURA, ADRIANA MIYAZAKI; LOPEZ PINAYA, WALTER HUGO; GADELHA, ARY; ZUGMAN, ANDRE; NOTO, CRISTIANO; CORDEIRO, QUIRINO; BELANGERO, SINTIA IOLE; JACKOWSKI, ANDREA P.; BRESSAN, RODRIGO A.; SATO, JOAO RICARDO. Investigating brain structural patterns in first episode psychosis and schizophrenia using MRI and a machine learning approach. PSYCHIATRY RESEARCH-NEUROIMAGING, v. 275, p. 14-20, MAY 30 2018. (13/10498-6, 13/08531-5, 14/07280-1, 13/05168-7)

VIEIRA, SANDRA; GONG, QI-YONG; PINAYA, WALTER H. L.; SCARPAZZA, CRISTINA; TOGNIN, STEFANIA; CRESPO-FACORRO, BENEDICTO; TORDESILLAS-GUTIERREZ, DIANA; ORTIZ-GARCIA, VICTOR; SETIEN-SUERO, ESTHER; SCHEEPERS, FLOORTJE E.; et al. Using Machine Learning and Structural Neuroimaging to Detect First Episode Psychosis: Reconsidering the Evidence. SCHIZOPHRENIA BULLETIN, v. 46, n. 1, p. 17-26, JAN 2020. (13/05168-7)

VIEIRA, SANDRA; PINAYA, WALTER H. L.; MECHELLI, ANDREA. Using deep learning to investigate the neuroimaging correlates of psychiatric and neurological disorders: Methods and applications. NEUROSCIENCE AND BIOBEHAVIORAL REVIEWS, v. 74, p. 18-pg., 2017-03-01. (13/05168-7)