Multimodal Machine Learning Algorithm as a Tool for Dementia Clinical Trial Screening

Dementia is a complex disease due to various etiologies. However, new multimodal machine learning algorithms were developed to improve the diagnosis of dementia into different categories, including normal cognition (NC), mild cognitive impairment (MCI), Alzheimer’s disease (AD), and non-AD dementias (nADD). One of the core difficulties in implementing dementia clinical trials, especially the AD trials lies in the diagnostic ambiguity of Alzheimer's, where symptomatic overlap with other cognitive disorders often leads to misdiagnosis. There are usually high screen failure rates in dementia clinical trials, which is a burden for the sponsor due to the manual verification of the screening that hinders the process. In this study, we aimed to explore the application of a multimodal machine learning algorithm as a tool for clinical trial patients’ disease screening verification to reduce the cost of the clinical study by improving its quality, accessibility, and speed. The accuracy assessment of the machine learning algorithm will be presented by comparing its results to a neurologist’s evaluation. The results will be described using the sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) in a real-world clinical trial setting. We will explore the optimal set of input variables used for the algorithm to balance the medical exams' accuracy, cost, and time.