Random forest prediction of Alzheimer’s disease using pairwise selection from time series data

2018/08/09 by Paul Moore, Paul J. Moore, Terry Lyons +2 · 1 citation
Biochemistry, Genetics and Molecular Biology · Mathematics · Medicine · Neuroscience · #Alzheimer's Disease Neuroimaging Initiative #Alzheimer's disease #Alzheimer's disease research and treatments #Artificial intelligence #Benchmark (surveying) #Computer science #Data mining #Dementia and Cognitive Impairment Research #Disease #Functional Brain Connectivity Studies #Machine learning #Mathematics #Medicine #Missing data #Pairwise comparison #Pathology #Pattern recognition (psychology) #Random forest #Statistics #Support vector machine #Time series #msc:62M10 #q-bio.QM #stat.AP

paper · pdf · doi:10.1371/journal.pone.0211558

6 pages, 1 figure, 6 tables

arxiv created 2018/08/09 · openalex publication_date 2019/02/14 · arxiv updated 2019/03/06 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/06

Abstract

Time-dependent data collected in studies of Alzheimer's disease usually has missing and irregularly sampled data points. For this reason time series methods which assume regular sampling cannot be applied directly to the data without a pre-processing step. In this paper we use a random forest to learn the relationship between pairs of data points at different time separations. The input vector is a summary of the time series history and it includes both demographic and non-time varying variables such as genetic data. To test the method we use data from the TADPOLE grand challenge, an initiative which aims to predict the evolution of subjects at risk of Alzheimer's disease using demographic, physical and cognitive input data. The task is to predict diagnosis, ADAS-13 score and normalised ventricles volume. While the competition proceeds, forecasting methods may be compared using a leaderboard dataset selected from the Alzheimer's Disease Neuroimaging Initiative (ADNI) and with standard metrics for measuring accuracy. For diagnosis, we find an mAUC of 0.82, and a classification accuracy of 0.73 compared with a benchmark SVM predictor which gives mAUC = 0.62 and BCA = 0.52. The results show that the method is effective and comparable with other methods.

Citations

• A Simple Generalisation of the Area Under the ROC Curve for Multiple Class Classification Problems
• TADPOLE Challenge: Prediction of Longitudinal Evolution in Alzheimer's Disease

Cited by

• Using path signatures to predict a diagnosis of Alzheimer's disease

Related

• Intelligent Diagnosis of Alzheimer's Disease Based on Machine Learning
• Forecasting the Progression of Alzheimer's Disease Using Neural Networks and a Novel Pre-Processing Algorithm
• Machine Learning Classification of Alzheimer's Disease Stages Using Cerebrospinal Fluid Biomarkers Alone
• Machine learning for the life-time risk prediction of Alzheimer’s disease: a systematic review
• Application of Machine Learning to Predict the Risk of Alzheimer's Disease: An Accurate and Practical Solution for Early Diagnostics