Predicting Longitudinal Outcomes of Alzheimer's Disease via a Tensor-based Joint Classification and Regression Model

Lodewijk Brand, Kai Nichols, Hua Wang, Heng Huang, Li Shen

PSB - 2020

Alzheimer's disease (AD) is a serious neurodegenerative condition that affects millions of people across the world. Recently, machine learning models have been used to predict the progression of AD, although they frequently do not take advantage of the longitudinal and structural components associated with multimodal medical data. Here, we present a new multi-block alternating direction method of multipliers algorithm to optimize the proposed _Joint Multi-Modal Longitudinal Regression and Classification_ objective. In our approach, we combine multimodal longitudinal clinical data of various modalities to simultaneously predict the cognitive scores and diagnoses of participants in the Alzheimer's Disease Neuroimaging Initiative. Our algorithm is designed to leverage thestructure associated with clinical data that is not incorporated into standard machine learning optimization algorithms. The approach shows state-of-the-art predictive performance and validates a collection of brain and genetic biomarkers that have been recorded previously in AD literature. In addition, our algorithm identifies twenty-one genetic biomarkers that have not previously been reported.


Cite this paper

MLA Copied to clipboard!
Brand, Lodewijk, et al. "Predicting Longitudinal Outcomes of Alzheimer’s Disease via a Tensor-Based Joint Classification and Regression Model." Pac Symp Biocomput. 2020.
BibTeX Copied to clipboard!
  title={Predicting Longitudinal Outcomes of Alzheimer's Disease via a Tensor-Based Joint Classification and Regression Model},
  author={Brand, Lodewijk and Nichols, Kai and Wang, Hua and Huang, Heng and Shen, Li},
  booktitle={Pac Symp Biocomput},
  organization={World Scientific}