The Promise of Machine Learning for Psychiatry

The abilities of movie-watching functional connectivity in individual identifications and individualized predictions

Quite a few studies have been performed based on movie-watching functional connectivity (FC). As compared to its resting-state counterpart, however, there is still much to know about its abilities in individual identifications and individualized predictions. To pave the way for appropriate usage of movie-watching FC, we systemically evaluated the minimum number of time points, as well as the exact functional networks, supporting individual identifications and individualized predictions of apparent traits based on it. We performed the study based on the 7T movie-watching fMRI data included in the HCP S1200 Release, and took IQ as the test case for the prediction analyses. The results indicate that movie-watching FC based on only 15 time points can support successful individual identifications (99.47%), and the connectivity contributed more to identifications were much associated with higher-order cognitive processes (the secondary visual network, the frontoparietal network and the posterior multimodal network). For individualized predictions of IQ, it was found that successful predictions necessitated 60 time points (predicted vs. actual IQ correlation significant at P < 0.05, based on 5,000 permutations), and the prediction accuracy increased logarithmically with the number of time points used for connectivity calculation. Furthermore, the connectivity that contributed more to individual identifications exhibited the strongest prediction ability. Collectively, our findings demonstrate that movie-watching FC can capture rich information about human brain function, and its ability in individualized predictions depends heavily on the length of fMRI scans.

Sampling inequalities affect generalization of neuroimaging-based diagnostic classifiers in psychiatry

BACKGROUND

The development of machine learning models for aiding in the diagnosis of mental disorder is recognized as a significant breakthrough in the field of psychiatry. However, clinical practice of such models remains a challenge, with poor generalizability being a major limitation.

METHODS

Here, we conducted a pre-registered meta-research assessment on neuroimaging-based models in the psychiatric literature, quantitatively examining global and regional sampling issues over recent decades, from a view that has been relatively underexplored. A total of 476 studies (n = 118,137) were included in the current assessment. Based on these findings, we built a comprehensive 5-star rating system to quantitatively evaluate the quality of existing machine learning models for psychiatric diagnoses.

RESULTS

A global sampling inequality in these models was revealed quantitatively (sampling Gini coefficient (G) = 0.81, p < .01), varying across different countries (regions) (e.g., China, G = 0.47; the USA, G = 0.58; Germany, G = 0.78; the UK, G = 0.87). Furthermore, the severity of this sampling inequality was significantly predicted by national economic levels (β =  - 2.75, p < .001, R²_adj = 0.40; r =  - .84, 95% CI: - .41 to - .97), and was plausibly predictable for model performance, with higher sampling inequality for reporting higher classification accuracy. Further analyses showed that lack of independent testing (84.24% of models, 95% CI: 81.0-87.5%), improper cross-validation (51.68% of models, 95% CI: 47.2-56.2%), and poor technical transparency (87.8% of models, 95% CI: 84.9-90.8%)/availability (80.88% of models, 95% CI: 77.3-84.4%) are prevailing in current diagnostic classifiers despite improvements over time. Relating to these observations, model performances were found decreased in studies with independent cross-country sampling validations (all p < .001, BF₁₀ > 15). In light of this, we proposed a purpose-built quantitative assessment checklist, which demonstrated that the overall ratings of these models increased by publication year but were negatively associated with model performance.

CONCLUSIONS

Together, improving sampling economic equality and hence the quality of machine learning models may be a crucial facet to plausibly translating neuroimaging-based diagnostic classifiers into clinical practice.

Enhanced Performance by Interpretable Low-Frequency Electroencephalogram Oscillations in the Machine Learning-Based Diagnosis of Post-traumatic Stress Disorder

Electroencephalography (EEG)-based diagnosis of psychiatric diseases using machine-learning approaches has made possible the objective diagnosis of various psychiatric diseases. The objective of this study was to improve the performance of a resting-state EEG-based computer-aided diagnosis (CAD) system to diagnose post-traumatic stress disorder (PTSD), by optimizing the frequency bands used to extract EEG features. We used eyes-closed resting-state EEG data recorded from 77 PTSD patients and 58 healthy controls (HC). Source-level power spectrum densities (PSDs) of the resting-state EEG data were extracted from 6 frequency bands (delta, theta, alpha, low-beta, high-beta, and gamma), and the PSD features of each frequency band and their combinations were independently used to discriminate PTSD and HC. The classification performance was evaluated using support vector machine with leave-one-out cross validation. The PSD features extracted from slower-frequency bands (delta and theta) showed significantly higher classification performance than those of relatively higher-frequency bands. The best classification performance was achieved when using delta PSD features (86.61%), which was significantly higher than that reported in a recent study by about 13%. The PSD features selected to obtain better classification performances could be explained from a neurophysiological point of view, demonstrating the promising potential to develop a clinically reliable EEG-based CAD system for PTSD diagnosis.

Bridging structural MRI with cognitive function for individual level classification of early psychosis via deep learning

INTRODUCTION

Recent efforts have been made to apply machine learning and deep learning approaches to the automated classification of schizophrenia using structural magnetic resonance imaging (sMRI) at the individual level. However, these approaches are less accurate on early psychosis (EP) since there are mild structural brain changes at early stage. As cognitive impairments is one main feature in psychosis, in this study we apply a multi-task deep learning framework using sMRI with inclusion of cognitive assessment to facilitate the classification of patients with EP from healthy individuals.

METHOD

Unlike previous studies, we used sMRI as the direct input to perform EP classifications and cognitive estimations. The proposed deep learning model does not require time-consuming volumetric or surface based analysis and can provide additionally cognition predictions. Experiments were conducted on an in-house data set with 77 subjects and a public ABCD HCP-EP data set with 164 subjects.

RESULTS

We achieved 74.9 ± 4.3% five-fold cross-validated accuracy and an area under the curve of 71.1 ± 4.1% on EP classification with the inclusion of cognitive estimations.

DISCUSSION

We reveal the feasibility of automated cognitive estimation using sMRI by deep learning models, and also demonstrate the implicit adoption of cognitive measures as additional information to facilitate EP classifications from healthy controls.

A machine learning case-control classifier for schizophrenia based on DNA methylation in blood

Epigenetic dysregulation is thought to contribute to the etiology of schizophrenia (SZ), but the cell type-specificity of DNA methylation makes population-based epigenetic studies of SZ challenging. To train an SZ case-control classifier based on DNA methylation in blood, therefore, we focused on human genomic regions of systemic interindividual epigenetic variation (CoRSIVs), a subset of which are represented on the Illumina Human Methylation 450K (HM450) array. HM450 DNA methylation data on whole blood of 414 SZ cases and 433 non-psychiatric controls were used as training data for a classification algorithm with built-in feature selection, sparse partial least squares discriminate analysis (SPLS-DA); application of SPLS-DA to HM450 data has not been previously reported. Using the first two SPLS-DA dimensions we calculated a "risk distance" to identify individuals with the highest probability of SZ. The model was then evaluated on an independent HM450 data set on 353 SZ cases and 322 non-psychiatric controls. Our CoRSIV-based model classified 303 individuals as cases with a positive predictive value (PPV) of 80%, far surpassing the performance of a model based on polygenic risk score (PRS). Importantly, risk distance (based on CoRSIV methylation) was not associated with medication use, arguing against reverse causality. Risk distance and PRS were positively correlated (Pearson r = 0.28, P = 1.28 × 10^-12), and mediational analysis suggested that genetic effects on SZ are partially mediated by altered methylation at CoRSIVs. Our results indicate two innate dimensions of SZ risk: one based on genetic, and the other on systemic epigenetic variants.