Machine learning infrastructure for Alzheimer’s disease research

Details

Supervisors

Macq, Benoît

Faculty

Ecole polytechnique de Louvain

Degree label

Master [120] : ingénieur civil en informatique, à finalité spécialisée

Abstract

Researchers have used increasingly advanced data analysis tools as they’ve studied the evolution of Alzheimer’s disease. This interest in data sciences led to new research merging machine learning and the medical field. Modern techniques and data storage increase the quantity of gathered data. It is interesting to see how machine learn- ing infrastructure may benefit medical research by handling, storing, and analyzing data. This paper’s major goal is to see how data management and analysis (statistical tools and machine learning) might be used in medical research, namely in the follow-up of Alzheimer’s patients. In order to build a powerful machine learning model, researchers must gather and concentrate data in a database. Machine learning will be used to predict a patient’s cognitive deterioration. The end-to-end procedure from data collection to machine learning output will represent the pipeline supplied to the research team as a final result to support them in their job. The pipeline is provided with a purpose of continuous learning, which can be used in order to address the issue regarding the possible lack of data. Despite modest predictive power at the time, machine learning in medical research combined with continuous learning seem very promising.