Applied machine learning in Alzheimer's disease research: omics, imaging, and clinical data.

Xiaoqian Jiang,Yizhuo Wang,Yejin Kim,Ziyi Li

doi:10.1042/etls20210249

Abstract

Alzheimer's disease (AD) remains a devastating neurodegenerative disease with few preventive or curative treatments available. Modern technology developments of high-throughput omics platforms and imaging equipment provide unprecedented opportunities to study the etiology and progression of this disease. Meanwhile, the vast amount of data from various modalities, such as genetics, proteomics, transcriptomics, and imaging, as well as clinical features impose great challenges in data integration and analysis. Machine learning (ML) methods offer novel techniques to address high dimensional data, integrate data from different sources, model the etiological and clinical heterogeneity, and discover new biomarkers. These directions have the potential to help us better manage the disease progression and develop novel treatment strategies. This mini-review paper summarizes different ML methods that have been applied to study AD using single-platform or multi-modal data. We review the current state of ML applications for five key directions of AD research: disease classification, drug repurposing, subtyping, progression prediction, and biomarker discovery. This summary provides insights about the current research status of ML-based AD research and highlights potential directions for future research.

Highlights

Alzheimer’s disease (AD) impacted more than five million Americans in 2020, which has imposed a huge psychological and economic burden on patients, their families, and society [1]
Drug Administration bringing new hope to those suffering from AD
Many past efforts have tried to delineate the number of AD distinct subtypes

Summary

Introduction

Alzheimer’s disease (AD) impacted more than five million Americans in 2020, which has imposed a huge psychological and economic burden on patients, their families, and society [1]. Some studies only used baseline information to predict the MCI-to-AD conversion in the future and the ML methods they adopted include support vector machine with linear kernel [117], multi-task neural network classifier [83], logistic regression [118] and multi-kernel learning [119] When longitudinal measurements, such as lab results and cognitive tests, were available, the problem of predicting disease progression was more complicated. While existing studies demonstrate that ML models using imaging data generally have higher diagnosis accuracy than using other data sources, the omics data types (e.g. genomics, metabolomics, as well epigenetics) provide a better chance to identify alterable gene/pathway markers or metabolite targets [84,143] This could be another promising direction for ML and DL to play an important role in future research. A user-friendly software allows more researchers and clinicians to apply the methods in practice

Conclusion

Summary