Machine learning is the study of algorithms and mathematical models that scientists rely on to progressively improve their performance on a specific task; these algorithms have the potential to provide a computational framework for predicting who will get Alzheimer’s disease.
The ability to identify those patients with memory impairment who are most likely to decline towards Alzheimer’s disease would be a valuable tool in early diagnosis. Predictions are complicated by the heterogeneity of the subject population as well as the complexity of clinical data that includes cognitive assessments, imaging, and MRI data, as well as genetic and demographic information. Machine learning could potentially provide a computational framework for predicting whom will get Alzheimer’s disease with the hope that the diagnosis would happen earlier.
Several researchers who receive funding from Cure Alzheimer’s Fund are engaged in research that incorporates computational techniques to understand the mechanistic basis of human disease. Manolis Kellis, for example, runs a laboratory at MIT that develops resources to study how genetic variation impacts gene expression and the implications for Alzheimer’s disease.
For more information on how Dr. Kellis is incorporating computational methods into his research, please see the link to his profile on our website.