Using deep belief network modelling to characterize differences in brain morphometry in schizophrenia

Abstract

Neuroimaging-based models contribute to increasing our understanding of schizophrenia pathophysiology and can reveal the underlying characteristics of this and other clinical conditions. However, the considerable variability in reported neuroimaging results mirrors the heterogeneity of the disorder. Machine learning methods capable of representing invariant features could circumvent this problem. In this structural MRI study, we trained a deep learning model known as deep belief network (DBN) to extract features from brain morphometry data and investigated its performance in discriminating between healthy controls (N = 83) and patients with schizophrenia (N = 143). We further analysed performance in classifying patients with a first-episode psychosis (N = 32). The DBN highlighted differences between classes, especially in the frontal, temporal, parietal, and insular cortices, and in some subcortical regions, including the corpus callosum, putamen, and cerebellum. The DBN was slightly more accurate as a classifier (accuracy = 73.6%) than the support vector machine (accuracy = 68.1%). Finally, the error rate of the DBN in classifying first-episode patients was 56.3%, indicating that the representations learned from patients with schizophrenia and healthy controls were not suitable to define these patients. Our data suggest that deep learning could improve our understanding of psychiatric disorders such as schizophrenia by improving neuromorphometric analyses.

Using deep autoencoders to identify abnormal brain structural patterns in neuropsychiatric disorders: A large‐scale multi‐sample study

“…In Salvador et al (2017), for example, the author's classified 128 patients with SCZ and 127 HC subjects using a number of structural features, including cortical volume and thickness; similar to our study, the use of SVM classifiers resulted in modest performance, with accuracies around 60%. In Pinaya et al (2016), using 143 patients with SCZ and 83 HC subjects, the SVM classifier achieved a balanced accuracy of 68.1%. Using 22 children with ASD and 16 HC subjects, Jiao et al (2010) were able to achieve an AUC-ROC of 0.93, however, the very low number of subjects may have inflated the estimate of performance (Schnack & Kahn, 2016).…”

Factors associated with opioid cessation: a machine learning approach

“…Several other studies used only regression-based methods to predict opioid and stimulant dependence (36), cocaine dependence(37) and alcohol dependence (38), which might not capture other relationships among variables. Several of the non-regression based methods we employed have also been applied in other studies, which focused mainly on MRI brain images as predictors of substance use disorder diagnoses (18,19,21,34).…”

“…Many statistical methods are limited in their ability to sort through large numbers of predictors (14). Data mining using machine learning, which is particularly well suited for identifying predictive factors among thousands of variables (15,16), has successfully identified predictor variables for a diverse set of outcomes (17)(18)(19)(20)(21). Here, we applied multiple machine learning techniques to evaluate a large set of clinical, demographic, general health, and behavioral variables in a large, racially mixed cohort who were recruited for genetic studies of substance dependence and not necessarily treated for opioids.…”

Precision Psychiatry Applications with Pharmacogenomics: Artificial Intelligence and Machine Learning Approaches

“…Moreover, deep belief networks have been used to characterize differences between healthy controls and young children with autism spectrum disorders based on functional MRI data [41]. Additionally, Pinaya et al [42] employed a deep belief network predictive model to differentiate healthy controls from patients with schizophrenia (accuracy = 73.6%) based on MRI data. Only MRI data and no multi-omics data were involved in their study.…”