Deep learning models for neuroimaging have largely been developed for individual tasks, limiting knowledge transfer across applications. Here we introduce GenFAR, a modular deep learning framework that learns general, clinically informed features from brain MRIs. We trained this modular architecture on 49,246 individuals across 11 cohorts, using 17 diverse classification and regression tasks spanning cognition, clinical, diagnosis, demographics, and biomarkers. This yields aggregated, focused feature sets that capture rich, clinically- and biologically-relevant brain representations. We developed a sequential learning approach where tasks progressively build on previously learned representations. Through an analysis of 5,000 task sequences, we identified an optimal sequence length of six tasks and introduced a Donor Score metric to quantify each task's contribution to downstream performance. This analysis revealed five consistently strong donor tasks (Age, AD/MCI, MMSE, Hypertension, Hyperlipidemia) that formed the base of our sequential model. We demonstrated the utility of our learned representation, in various tasks beyond those included in the training set, to serve as the foundation for specialized secondary predictors. We further showed that using the learned feature representation can substantially increase the sample efficiency of secondary deep learning training tasks and models, as well as improve their accuracy.
V. Bashyam, G. Erus, Junhao Wen et al.· 0 citations
Leveraging multi-omics to better understand the molecular signatures and pathways underlying Alzheimer’s disease (AD) pathogenesis is critical for early diagnosis and disease modifying interventions. We performed peripheral blood transcriptome (N = 669) and epigenome microarray analyses (N = 553) on non-Hispanic white participants from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) to identify molecular signatures of AD. We identified specific transcripts (e.g., MAPK14, GM2A, CD177) and co-expression networks that were dysregulated in AD, marked by a strong influence of APOE ε4 genotype, and characterized by a consistent pattern of immune activation, inflammation, and metabolic suppression. Further, these peripheral signatures were linked to central AD pathology (amyloid PET, CSF p-tau181) and neurodegeneration (plasma NfL, regional atrophy), with two genes, MXD3 and NR4A1, identified as protective against progression from MCI to AD. Our work emphasizes the importance of APOE genotypes in AD pathophysiology and highlights potential targets for biomarker discovery and personalized therapeutic strategies.
Brendan A. Mitchell, I. Hausle, Sarah Smith et al.· npj Dementia· 0 citations
While most studies of Alzheimer's disease (AD) examine cross‐sectional relationships among biomarkers, longitudinal relationships are also highly relevant.
B. Saef, K. K. Petersen, Katherine E. Volluz et al.· Alzheimer's & Dementia· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
