Classification of tau status with machine learning models in amyloid-positive cohorts.
Authors
Affiliations (4)
Affiliations (4)
- Department of Biomedical Engineering, School of Engineering, University of Alabama at Birmingham, Birmingham, Alabama, USA.
- Alzheimer's Disease Research Center, Heersink School of Medicine, University of Alabama at Birmingham, Birmingham, Alabama, USA.
- Department of Radiology, Heersink School of Medicine, University of Alabama at Birmingham, Birmingham, Alabama, USA.
- Department of Neurology, Heersink School of Medicine, University of Alabama at Birmingham, Birmingham, Alabama, USA.
Abstract
Although tau positron emission tomography (PET) imaging is effective for staging tau pathology, it is limited clinically by cost and availability. Machine learning models based on magnetic resonance imaging (MRI)- and amyloid PET-derived features may serve as useful screening tools for tau pathology. Multiple machine learning models were developed to classify tau positivity in the Braak III/IV region using structural MRI, amyloid PET, and demographic features. Alzheimer's Disease Neuroimaging Initiative (ADNI) (n = 410) data were used for model training. Open Access Series of Imaging Studies (OASIS-3; n = 143) and the Standardized Centralized Alzheimer's Disease Neuroimaging (SCAN; n = 154) data were used for external validation. Logistic regression achieved the best performance with areas under the curve (AUCs) of 0.92 for both internal and external validation. Combined external validation yielded accuracy/sensitivity/specificity of 85%/83%/85%. Subjects with mild cognitive impairment and predicted tau positivity progressed to AD at a significantly faster pace (p < 10<sup>-6</sup>). Our model demonstrates the feasibility of classifying tau burden in amyloid-positive cohorts with MRI- and amyloid PET-derived features and may serve as a surrogate biomarker.