AUTHOR=Salvador Raymond , Canales-Rodríguez Erick , Guerrero-Pedraza Amalia , Sarró Salvador , Tordesillas-Gutiérrez Diana , Maristany Teresa , Crespo-Facorro Benedicto , McKenna Peter , Pomarol-Clotet Edith 

TITLE=Multimodal Integration of Brain Images for MRI-Based Diagnosis in Schizophrenia

JOURNAL=Frontiers in Neuroscience

VOLUME=Volume 13 - 2019

YEAR=2019

URL=https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2019.01203

DOI=10.3389/fnins.2019.01203

ISSN=1662-453X

ABSTRACT=Magnetic resonance imaging (MRI) has been proposed as a source of information for automatic prediction of individual diagnostic in schizophrenia. Optimal integration of data from different MRI modalities is an active area of research aimed at increasing diagnostic accuracy. Based on a sample of 96 patients with schizophrenia and a matched sample of 115 healthy controls that had undergone a single multimodal MRI session we generated individual brain maps of grey matter vbm, 1back and 2back levels of activation (nback fMRI), maps of amplitude of low frequency fluctuations (resting fMRI) and maps of weighted global brain connectivity (resting fMRI). Four unimodal classifiers (Ridge, Lasso, Random Forests and Gradient boosting) were applied to these maps to evaluate their classification accuracies. Based on the assignments made by the algorithms on test individuals we quantified the amount of predictive information shared between maps (what we call redundancy analysis). Finally, we explored the added accuracy provided by a set of multimodal strategies that included post-classification integration based on probabilities, two step sequential integration and voxel level multimodal integration through 1D-convolutional neural networks (1D-CNN). All four unimodal classifiers showed the highest test accuracies with the 2back maps (80% on average) achieving a maximum of 84% with the Lasso. Redundancy levels between brain maps were generally low (overall mean redundancy score of 0.14 in a 0-1 range) indicating that each brain map contained differential predictive information. The highest multimodal accuracy was delivered by the two step Ridge classifier (87%) followed by the Ridge maximum and mean probability classifiers (both with 85% accuracy) and by the 1D-CNN that achieved the same accuracy than the best unimodal classifier (84%). From the results we conclude that from all MRI modalities evaluated, task based fMRI may be the best option for unimodal diagnostic in schizophrenia. Low redundancy values point to ample potential for accuracy improvements through multimodal integration, with the two step Ridge being a suitable strategy.