Digital Biomarkers

Multi-sensor Disease Detection Algorithms

Background

Early detection and monitoring of Alzheimer’s are imperative for the successful implementation of treatments that may reverse the disease. We need systems that can passively and continuously screen as many patients as possible, and identify those at higher risk.

Approach

Circadic created and implemented a symptom map using a multitude of data from smartphone and smartwatch sensors to reveal those at higher risk for pathologic cognitive decline.

Under the hood

Circadic helped Tech and Pharma partners create algorithms to compile a digital phenotype of the disease. The system was used in a Pharma-led pilot study that collected cross sectional data amongst healthy controls and mild cognitively impaired participants. Sensor data was used as a training set, diagnosis was used as ground truth. The algorithm was capable of discriminating between cognitively normal and mildly impaired subjects with an AUC=0.82.

‍Kourtis, L.C. et al. Digital biomarkers for Alzheimer’s disease: the mobile/wearable devices opportunity. npj Digital Med 2, 9 (2019)