Improving Alzheimer’s Trial Success Through AI

This scientific discussion will review how emerging AI solutions that employ quantitative neuroimaging can help address some of the key challenges faced in Alzheimer’s Disease (AD) clinical trials including:

Improved stratification

One important reason that AD trials have failed to show positive results is the inherent heterogeneity of the AD population. This has resulted in trials being underpowered due to the inclusion of patients who were not going to decline during the trial and thus were not able to demonstrate a treatment effect.

We will discuss the recent focus on improving stratification by using machine learning to predict a patient’s likely risk to decline over the period of the trial and the impact of using these models to enrich patient selection.

Better imaging outcomes

We will present techniques that use deep learning for quantitative segmentation of structural brain regions and the added value they provide over purely qualitative assessments.

We will also provide an overview of the AD trials that have employed quantitative imaging as an endpoint and why providing these measures is important.

Safety and prescribing

Given the controversies around anti-amyloid drug side effects (ARIA), we will discuss the potential for AI methods to detect microbleeds as well as to support a personalized medicine approach.