External validation of an MRI‐based classifier of arteriolar sclerosis
Alzheimer's & Dementia(2020)
Abstract
Background Arteriolar sclerosis is a common age‐related neuropathology that has been associated with lower cognition and higher risk of dementia. Definitive diagnosis of arteriolar sclerosis is only possible at autopsy. We recently introduced an MRI‐based classifier of arteriolar sclerosis [1]. In the present work, we tested this classifier for prediction of pathology in‐vivo, and also validated it on external datasets. Method In‐vivo 3T MRI and pathology data were available on participants of the Rush Memory and Aging Project (MAP) and Religious Orders Study (ROS), two longitudinal cohort studies of aging used in training of the classifier. We first assessed the performance of the classifier in predicting arteriolar sclerosis based on in‐vivo MRI of MAP and ROS participants that had not been included in the training sample. The generalizability of the classifier was then evaluated in the Minority Aging Research Study (MARS), and Alzheimer's Disease Neuroimaging Initiative (ADNI), by testing the association of the classification confidence score with change in cognition two years after baseline MRI, using Spearman’s rank correlation. Result The area under the receiver operating characteristic curve for in‐vivo classification of arteriolar sclerosis in MAP and ROS data not used in training of the classifier was 0.77. Also, the classification confidence score showed significant association with two‐year decline in perceptual speed (p=0.027) in the MARS dataset, and with Trail Making Test Part A (p=0.027), and Trail Making Test Part B (p=0.046) in the ADNI dataset. Conclusion Our previously introduced classifier of arteriolar sclerosis was shown to have a high performance in prediction of the pathology based on in‐vivo MRI, illustrating its significance. Testing in MARS and ADNI showed that the classifier generalizes to other cohorts and demonstrated its clinical utility in predicting cognitive decline two years after baseline MRI. References: [1] N. Makkinejad, A.M. Evia, A. Tamhane, D.A. Bennett, J.A. Schneider, K. Arfanakis, A Novel MRI Classifier of Arteriolar Sclerosis in Aging: Prediction of Pathology and Cognitive Decline, Proc. Int. Soc. for Magn. Reson. in Med. 2019.
MoreTranslated text
Key words
mri‐based,classifier,external validation
AI Read Science
Must-Reading Tree
Example
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
Generate MRT to find the research sequence of this paper
Chat Paper
Summary is being generated by the instructions you defined