Leveraging Mobile Sensing and Machine Learning for Personalized Mental Health Care

Abstract

Mental illness is widespread in our society, yet remains difficult to treat due to challenges such as stigma and overburdened health care systems. New paradigms are needed for treating mental illness outside the practitioner’s office. We propose a framework to guide the design of mobile sensing systems for personalized mental health interventions. This framework guides researchers in constructing interventions from the ground up through four phases: sensor data collection, digital biomarker extraction, health state detection, and intervention deployment. We highlight how this framework advances research in personalized mHealth and address remaining challenges, such as ground truth fidelity and missing data.