한빛사
>
한빛사논문
한빛사논문
- 조회121
Alzheimers. Dement., Jul 2025, 21 (7) | https://doi.org/10.1002/alz.70362
Differential dementia detection from multimodal brain images in a real-world dataset
Matthew Leming1,2 Hyungsoon Im1,2,3
1Center for Systems Biology, Massachusetts General Hospital, Boston, USA
2Massachusetts Alzheimer’s Disease Research Center, Massachusetts General Hospital, Boston, USA
3Department of Radiology, Massachusetts General Hospital, Boston, USA
Correspondence : Matthew Leming and Hyungsoon Im
1Center for Systems Biology, Massachusetts General Hospital, Boston, USA
2Massachusetts Alzheimer’s Disease Research Center, Massachusetts General Hospital, Boston, USA
3Department of Radiology, Massachusetts General Hospital, Boston, USA
Correspondence : Matthew Leming and Hyungsoon Im
Abstract
Introduction: Artificial intelligence (AI) models have been applied to differential dementia detection tasks in brain images from curated, high-quality benchmark databases, but not real-world data in hospitals.
Methods: We describe a deep learning model specially trained for disease detection in heterogeneous clinical images from electronic health records without focusing on confounding factors. It encodes up to 14 multimodal images, alongside age and demographics, and outputs the likelihood of vascular dementia, Alzheimer's, Lewy body dementia, Pick's disease, mild cognitive impairment, and unspecified dementia. We use data from Massachusetts General Hospital (183,018 images from 11,015 patients) for training and external data (125,493 images from 6,662 patients) for testing.
Results: Performance ranged between 0.82 and 0.94 area under the curve (AUC) on data from 1003 sites.
Discussion: Analysis shows that the model focused on subcortical brain structures as the basis for its decisions. By detecting biomarkers in real-world data, the presented techniques will help with clinical translation of disease detection AI.
Highlights: Our artificial intelligence (AI) model can detect neurodegenerative disorders in brain imaging electronic health record (EHR) data. It encodes up to 14 brain images and text information from a single patient's EHR. Attention maps show that the model focuses on subcortical brain structures. Performance ranged from 0.82 to 0.94 area under the curve (AUC) on data from 1003 external sites.
Methods: We describe a deep learning model specially trained for disease detection in heterogeneous clinical images from electronic health records without focusing on confounding factors. It encodes up to 14 multimodal images, alongside age and demographics, and outputs the likelihood of vascular dementia, Alzheimer's, Lewy body dementia, Pick's disease, mild cognitive impairment, and unspecified dementia. We use data from Massachusetts General Hospital (183,018 images from 11,015 patients) for training and external data (125,493 images from 6,662 patients) for testing.
Results: Performance ranged between 0.82 and 0.94 area under the curve (AUC) on data from 1003 sites.
Discussion: Analysis shows that the model focused on subcortical brain structures as the basis for its decisions. By detecting biomarkers in real-world data, the presented techniques will help with clinical translation of disease detection AI.
Highlights: Our artificial intelligence (AI) model can detect neurodegenerative disorders in brain imaging electronic health record (EHR) data. It encodes up to 14 brain images and text information from a single patient's EHR. Attention maps show that the model focuses on subcortical brain structures. Performance ranged from 0.82 to 0.94 area under the curve (AUC) on data from 1003 external sites.
논문정보
- Research article
- 2025년 07월 (BRIC 등록일 2025-07-04)
- 국외 연구진
- 의약학 > 신경의학
댓글 작성 로그인
팝업 닫기관련 링크
연구자 키워드
관련분야 연구자보기
소속기관 논문보기
관련분야 논문보기
해당논문 저자보기
