HYDRA: Revealing heterogeneity of imaging and genetic patterns through a multiple max-margin discriminative analysis framework

详细信息    查看全文

• 作者：Erdem Varol ; ^{erdem.varol@uphs.upenn.edu} ; Aristeidis Sotiras ; Christos Davatzikos ; for the Alzheimer's Disease Neuroimaging Initiative¹
• 关键词：Semi-supervised pattern analysis ; Multivariate ; Max-margin classification ; Convex polytope ; SVM ; Clustering ; MRI ; Genetics ; Neuroimaging ; Heterogeneity ; Aging ; Alzheimer's disease ; ADNI
• 刊名：NeuroImage
• 出版年：2017
• 出版时间：15 January 2017
• 年：2017
• 卷：145
• 期：part_PB
• 页码：346-364
• 全文大小：2430 K
• 卷排序：145

文摘

We present a novel machine learning framework for the analysis of heterogeneity in neuroimaging studies. We propose a semi-supervised learning framework that integrates classification and clustering. The anatomical and genetic heterogeneity of Alzheimer's disease is explored using the proposed framework. The anatomical and genetic subtypes that are revealed are clinically meaningful and match well with previous studies.