Self-supervised multimodal learning for group inferences from MRI data: Discovering disorder-relevant brain regions and multimodal links.

In recent years, deep learning approaches have gained significant attention in predicting brain disorders using neuroimaging data. However, conventional methods often rely on single-modality data and supervised models, which provide only a limited perspective of the intricacies of the highly complex brain. Moreover, the scarcity of accurate diagnostic labels in clinical settings hinders the applicability of the supervised models. To address these limitations, we propose a novel self-supervised framework for extracting multiple representations from multimodal neuroimaging data to enhance group inferences and enable analysis without resorting to labeled data during pre-training. Our approach leverages Deep InfoMax (DIM), a self-supervised methodology renowned for its efficacy in learning representations by estimating mutual information without the need for explicit labels. While DIM has shown promise in predicting brain disorders from single-modality MRI data, its potential for multimodal data remains untapped. This work extends DIM to multimodal neuroimaging data, allowing us to identify disorder-relevant brain regions and explore multimodal links. We present compelling evidence of the efficacy of our multimodal DIM analysis in uncovering disorder-relevant brain regions, including the hippocampus, caudate, insula, - and multimodal links with the thalamus, precuneus, and subthalamus hypothalamus. Our self-supervised representations demonstrate promising capabilities in predicting the presence of brain disorders across a spectrum of Alzheimer's phenotypes. Comparative evaluations against state-of-the-art unsupervised methods based on autoencoders, canonical correlation analysis, and supervised models highlight the superiority of our proposed method in achieving improved classification performance, capturing joint information, and interpretability capabilities. The computational efficiency of the decoder-free strategy enhances its practical utility, as it saves compute resources without compromising performance. This work offers a significant step forward in addressing the challenge of understanding multimodal links in complex brain disorders, with potential applications in neuroimaging research and clinical diagnosis.

Neuroimaging correlates of Alzheimer's disease biomarker concentrations in a racially diverse high-risk cohort of middle-aged adults.

INTRODUCTION: In this study, we investigated biomarkers in a midlife, racially diverse, at-risk cohort to facilitate early identification and intervention. We examined neuroimaging measures, including resting state functional magnetic resonance imaging (fMRI), white matter hyperintensity vo (WMH), and hippocampal volumes, alongside cerebrospinal fluid (CSF) markers. METHODS: Our data set included 76 cognitively unimpaired, middle-aged, Black Americans (N = 29, F/M = 17/12) and Non-Hispanic White (N = 47, F/M = 27/20) individuals. We compared cerebrospinal fluid phosphorylated tau141 and amyloid beta (Aß)42 to fMRI default mode network (DMN) subnetwork connectivity, WMH volumes, and hippocampal volumes. RESULTS: Results revealed a significant race × Aß42 interaction in Black Americans: lower Aß42 was associated with reduced DMN connectivity and increased WMH volumes regions but not in non-Hispanic White individuals. DISCUSSION: Our findings suggest that precuneus DMN connectivity and temporal WMHs may be linked to Alzheimer's disease risk pathology during middle age, particularly in Black Americans. HIGHLIGHTS: Cerebrospinal fluid (CSF) amyloid beta (Aß)42 relates to precuneus functional connectivity in Black, but not White, Americans. Higher white matter hyperintensity volume relates to lower CSF Aß42 in Black Americans. Precuneus may be a hub for early Alzheimer's disease pathology changes detected by functional connectivity.

Depression, Vascular Burden, and Dementia Prevalence in Late Middle-Aged and Older Black Adults.

OBJECTIVES: Late-life depression and white matter hyperintensities (WMH) have been linked to increased dementia risk. However, there is a dearth of literature examining these relationships in Black adults. We investigated whether depression or WMH volume are associated with a higher likelihood of dementia diagnosis in a sample of late middle-aged to older Black adults, and whether dementia prevalence is highest in individuals with both depression and higher WMH volume. METHODS: Secondary data analysis involved 443 Black participants aged 55+ with brain imaging within 1 year of baseline visit in the National Alzheimer's Coordinating Center Uniform Data Set. Chi-square analyses and logistic regression models controlling for demographic variables examined whether active depression in the past 2 years, WMH volume, or their combination were associated with higher odds of all-cause dementia. RESULTS: Depression and higher WMH volume were associated with a higher prevalence of dementia. These associations remained after controlling for demographic factors, as well as vascular disease burden. Dementia risk was highest in the depression/high WMH volume group compared to the depression-only group, high WMH volume-only group, and the no depression/low WMH volume group. Post hoc analyses comparing the Black sample to a demographically matched non-Hispanic White sample showed associations of depression and the combination of depression and higher WMH burden with dementia were greater in Black compared to non-Hispanic White individuals. DISCUSSION: Results suggest late-life depression and WMH have independent and joint relationships with dementia and that Black individuals may be particularly at risk due to these factors.

A deep learning approach for mental health quality prediction using functional network connectivity and assessment data.

While one can characterize mental health using questionnaires, such tools do not provide direct insight into the underlying biology. By linking approaches that visualize brain activity to questionnaires in the context of individualized prediction, we can gain new insights into the biology and behavioral aspects of brain health. Resting-state fMRI (rs-fMRI) can be used to identify biomarkers of these conditions and study patterns of abnormal connectivity. In this work, we estimate mental health quality for individual participants using static functional network connectivity (sFNC) data from rs-fMRI. The deep learning model uses the sFNC data as input to predict four categories of mental health quality and visualize the neural patterns indicative of each group. We used guided gradient class activation maps (guided Grad-CAM) to identify the most discriminative sFNC patterns. The effectiveness of this model was validated using the UK Biobank dataset, in which we showed that our approach outperformed four alternative models by 4-18% accuracy. The proposed model's performance evaluation yielded a classification accuracy of 76%, 78%, 88%, and 98% for the excellent, good, fair, and poor mental health categories, with poor mental health accuracy being the highest. The findings show distinct sFNC patterns across each group. The patterns associated with excellent mental health consist of the cerebellar-subcortical regions, whereas the most prominent areas in the poor mental health category are in the sensorimotor and visual domains. Thus the combination of rs-fMRI and deep learning opens a promising path for developing a comprehensive framework to evaluate and measure mental health. Moreover, this approach had the potential to guide the development of personalized interventions and enable the monitoring of treatment response. Overall this highlights the crucial role of advanced imaging modalities and deep learning algorithms in advancing our understanding and management of mental health.