Visual deep learning of unprocessed neuroimaging characterises dementia subtypes and generalises across non-stereotypic samples
Year: 2023
Type: article
Abstract: Background Dementia's diagnostic protocols are mostly based on standardised neuroimaging data collected in the Global North from homogeneous samples. In other non-stereotypical samples (participants with diverse admixture, genetics, demographics, MRI signals, or cultural origins), classifications of disease are difficult due to demographic and region-specific sample heterogeneities, lower quality scanners, and non-harmonised pipelines. Methods We implemented a fully automatic computer-vision classifier using deep learning neural networks. A DenseNet was applied on raw (unpreprocessed) data from 3000 participants (behavioural variant frontotemporal dementia-bvFTD, Alzheimer's disease-AD, and healthy controls; both male and female as self-reported by participants). We tested our results in demographically matched and unmatched samples to discard possible biases and performed multiple out-of-sample validations. Findings Robust classification results across all groups were achieved from standardised 3T neuroimaging data from the Global North, which also generalised to standardised 3T neuroimaging data from Latin America. Moreover, DenseNet also generalised to non-standardised, routine 1.5T clinical images from Latin America. These generalisations were robust in samples with heterogenous MRI recordings and were not confounded by demographics (i.e., were robust in both matched and unmatched samples, and when incorporating demographic variables in a multifeatured model). Model interpretability analysis using occlusion sensitivity evidenced core pathophysiological regions for each disease (mainly the hippocampus in AD, and the insula in bvFTD) demonstrating biological specificity and plausibility. Interpretation The generalisable approach outlined here could be used in the future to aid clinician decision-making in diverse samples. Funding The specific funding of this article is provided in the acknowledgements section.
Source: EBioMedicine
Institutions: Harvard University, University of California, San Francisco, Trinity College Dublin, Adolfo Ibáñez University, University of San Andrés, Massachusetts General Hospital, Global Brain Health Institute, Erasmus MC, Consejo Nacional de Investigaciones Científicas y Técnicas, Universidad de Buenos Aires
Language: English
Subfield: Psychiatry and Mental health
Field: Medicine
Domain: Health Sciences
Open Access status: gold
Estimated APC paid: $5,000
Funders: Alzheimer's Association, Global Brain Health Institute, Alzheimer's Society, Consejo Nacional de Investigaciones Científicas y Técnicas, Agencia Nacional de Investigación y Desarrollo, National Institutes of Health, National Institute on Aging, Fondo para la Investigación Científica y Tecnológica, Fondo Nacional de Desarrollo Científico y Tecnológico, Fondo de Fomento al Desarrollo Científico y Tecnológico, Fondo de Financiamiento de Centros de Investigación en Áreas Prioritarias
Awards: G1011791342, G103627852, G1080512434, Multi-partner Consortium for Dementia Research in Latin America, G1371425577, G150854395, G1581799926, G1919018801, G1968784004, G2103439199, G2118524674, G2363475639, A novel deep learning algorithm for classification of dementia subtypes, G2574436759, MACHINE LEARNING FOR MENTAL HEALTH NEW DEVELOPMENTS AND APPLICATIONS USING MULTIMODAL DATA, G2822108074, G3206389968, G3626066913, G3628792472, G371657172, G3746078142, CENTRO DE GEROCIENCIA SALUD MENTAL Y METABOLISMO (GERO), G4216404362, G4255270288, G4914364266, G4921574523, G497235149, G5032928781, G5163463781, G5250673234, G546164382, MACHINE LEARNING PARA EL DIAGNOSTICO PRECOZ Y PERFIL MULTIMODAL DE LA ENFERMEDAD DE ALZHEIMER BASADO EN MIRNA DE EXOSOMAS CIRCULANTES., G6124283843, G6318961629, NEUREKA LATAM APLICACION MOVIL DE EVALUACION NEUROCOGNITIVA AUTOADMINISTRABLE PARA EL DIAGNOSTICO DE LA DEMENCIA EN POBLACION CHILENA DE ADULTOS MAYORES, SENSITIVE NEUROCOGNITIVE MARKERS OF NEURODEGENERATION A MULTI-FEATURE FRAMEWORK, G7007719478, G7049503461, G734147807, G7574155280, G7616641907, NATURAL SPEECH MARKERS OF NEURODEGENERATION AN INTERDISCIPLINARY APPROACH FOR DISEASE DISCRIMINATION AND SYMPTOM TRACKING, MAPPING AND MODELING THE HEALTHY AND PATHOLOGICAL AGING TRAJECTORIES OF THE HUMAN BRAIN, G883642430, US-South American Initiative for Genetic-Neural-Behavioral Interactions in Human Neurodegenerative Research
OpenAlex ID: W4360966363