Evaluation of the Alzheimer’s Disease Progression Score

[摘要] The Alzheimer’s Disease (AD) Progression Score (ADPS) method (Jedynak et al., Neuroimage 2012) was presented as an unsupervised statistical method which combines a heterogeneous collections of measurements of AD. It allows (1) the visualization of the cascade of mea- surements and (2) the assignment of an AD progression score (ADPS) to each subject and each time-point in a cohort. In this paper, we use the AD Neuroimaging Ini- tiative (ADNI) cohort to evaluate the ADPS using three standard benchmarks: (a) the prediction of the transition from Mild Cognitive Impairment (MCI) to AD dementia, (b) the sample size obtained for a simulated drug trial, and (c) the Cox proportional hazard model. We find that in all three benchmarks the ADPS method provides a progression score that is better than the best of the indi- vidual measurements available in ADNI.

[发布机构] Johns Hopkins University

[效力级别] s Disease

[关键词] Alzheimer"s Disease;Statistical Evaluation;Machine Learning;Mathematics