HOPES: An Integrative Digital Phenotyping Platform for Data Collection, Monitoring, and Machine Learning.

Xuancong Wang,Wijaya Martanto,Nikola Vouk,Robert Jt Morris,Creighton Heaukulani,Thisum Buddhika,Jimmy Lee

doi:10.2196/23984

Abstract

The collection of data from a personal digital device to characterize current health conditions and behaviors that determine how an individual’s health will evolve has been called digital phenotyping. In this paper, we describe the development of and early experiences with a comprehensive digital phenotyping platform: Health Outcomes through Positive Engagement and Self-Empowerment (HOPES). HOPES is based on the open-source Beiwe platform but adds a wider range of data collection, including the integration of wearable devices and further sensor collection from smartphones. Requirements were partly derived from a concurrent clinical trial for schizophrenia that required the development of significant capabilities in HOPES for security, privacy, ease of use, and scalability, based on a careful combination of public cloud and on-premises operation. We describe new data pipelines to clean, process, present, and analyze data. This includes a set of dashboards customized to the needs of research study operations and clinical care. A test use case for HOPES was described by analyzing the digital behavior of 22 participants during the SARS-CoV-2 pandemic.

Highlights

We are at an age in health care where we have much data at our disposal, including the high penetration of digital electronic medical records and advanced techniques available for their analysis [1]
In the section on The Health Outcomes through Positive Engagement and Self-Empowerment (HOPES) Platform and Its First Use in the HOPE-S Study, we describe the overall architecture of the HOPES platform
In the section Conclusions, we provide some overall conclusions that can be drawn from our experiences with digital phenotyping

Summary

Introduction

We are at an age in health care where we have much data at our disposal, including the high penetration of digital electronic medical records and advanced techniques available for their analysis [1]. It has been argued that lifestyle choices exceed the impact of health care received as a determinant of premature death [3] It has been highlighted by Onnela [4] that the wide adoption of smartphones and the increasing use of wearable devices open up a new vista of characterizing both current health conditions and the ongoing behaviors that will determine how an individual’s health will evolve. As examples of these new data sources, we can readily measure physical activity, heart rate, heart rate variability, temperature, sleep, sociability (amount of human interaction), and smartphone usage (amount and duration of use, type of use, and the way a screen is tapped and scrolled). The use of digital phenotyping both complements and extends the use of traditional https://www.jmir.org/2021/3/e23984

Objectives

Conclusion