Multimodal deep learning for Alzheimer’s disease dementia assessment
Мультимодальное глубокое обучение для оценки деменции при болезни Альцгеймера
2022-06-20
SCID: 54.1/8w9wdtvu
Discuss with AI
Alzheimer’s diseasedementia assessmentinterpretable computer visionmild cognitive impairmentmultimodal deep learning
Figures from the paper
Abstract (AI)
Worldwide, there are nearly 10 million new cases of dementia annually, of which Alzheimer's disease (AD) is the most common. New measures are needed to improve the diagnosis of individuals with cognitive impairment due to various etiologies. Here, we report a deep learning framework that accomplishes multiple diagnostic steps in successive fashion to identify persons with normal cognition (NC), mild cognitive impairment (MCI), AD, and non-AD dementias (nADD). We demonstrate a range of models capable of accepting flexible combinations of routinely collected clinical information, including demographics, medical history, neuropsychological testing, neuroimaging, and functional assessments. We then show that these frameworks compare favorably with the diagnostic accuracy of practicing neurologists and neuroradiologists. Lastly, we apply interpretability methods in computer vision to show that disease-specific patterns detected by our models track distinct patterns of degenerative changes throughout the brain and correspond closely with the presence of neuropathological lesions on autopsy. Our work demonstrates methodologies for validating computational predictions with established standards of medical diagnosis.
Key Findings
1
A successive deep learning framework distinguishes normal cognition, mild cognitive impairment, Alzheimer’s disease, and non-Alzheimer’s dementias.
2
Computer-vision interpretability reveals disease-specific brain degeneration patterns that correspond closely to neuropathological lesions observed at autopsy.
3
The frameworks achieve diagnostic accuracy that compares favorably with practicing neurologists and neuroradiologists.
4
The models accept flexible combinations of routinely collected demographics, medical history, neuropsychological tests, neuroimaging, and functional assessments.
5
The study demonstrates methods for validating computational diagnostic predictions against established medical diagnostic standards.
Research Object
Individuals with normal cognition, mild cognitive impairment, Alzheimer’s disease, and non-Alzheimer’s dementias
Research Subject
Multimodal diagnostic assessment and disease-specific neurodegenerative pattern detection across these cognitive and dementia conditions
Publication Details
Publication Date
2022-06-20
Journal
Publisher
ISSN
Cited by
512
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
References available in scid.ai3
A Unified Approach to Interpreting Model Predictions2017
The diagnosis of dementia due to Alzheimer's disease: Recommendations from the National Institute on Aging‐Alzheimer's Association workgroups on diagnostic guidelines for Alzheimer's disease2011
A Machine Learning Approach for the Differential Diagnosis of Alzheimer and Vascular Dementia Fed by MRI Selected Features2020