AI Predicts the Start of Mental Illness with Fitbit Wearable

Source: Geralt/Pixabay

A new mental health study published in Frontiers in Digital Health used artificial intelligence (AI) machine learning with biometric data from wearables and medical exams to predict the onset of mental illness.

According to Gallup, in the decade prior to the COVID-19 pandemic, mental health issues of sadness, stress, worry, and anger have been rising worldwide.

Our World in Data estimated that 792 million people lived with a mental health disorder, representing 10.7 percent, or over one in ten people worldwide in 2017. In America, 41.5 percent of adults exhibited symptoms of anxiety or depression in early 2021, according to the U.S. Centers for Disease Control and Prevention (CDC).

“Prevention and detection at the onset of mental illness are extremely important because of the generally low remission rate for mental illness and the favorable prognosis of early initiation of treatment,” wrote the researchers affiliated with The University of Tokyo and JMDC Inc. located in Tokyo, Japan.

The study used data from a JMDC Inc. health insurance database with over 4,600 subjects. The AI algorithm was fed three months of continuous wearable data and any medical exam data. The wearable data included information from Fitbit on activity, sleep, and resting heart rate.

“In this study, we used machine learning to build a predictive model that used sleep and activity data acquired from Fitbit wearable devices and medical examination records to establish the criteria leading to mental illness onset,” the researchers wrote.

The scientists created a predictive model using AI machine learning, specifically an XGBoost binary classification model. XGBoost, short for Extreme Gradient Boosting, is an open-source distributed gradient-boosted decision tree (GBDT) machine learning library. Gradient boosting is a supervised learning AI algorithm often used in regression and classification tasks that gives robust predictions.

The study showed that early intervention targeting the stabilization of sleep is an effective measure for the onset of mental illness. Sleep abnormalities were identified three months before the emergence of mental illness from the Fitbit wearable data.

“Analysis of the results of the model built with machine learning suggested that sleep abnormalities, especially the destabilization of sleep rhythms, are associated with an increased illness onset probability and that sleep disturbances may be a predictor of mental illness onset,” the researchers discovered. “Furthermore, activity-related indices and medical examination data relating to alcohol consumption were included in the topmost features, and these were also suggested to be factors.”

Copyright © 2022 Cami Rosso. All rights reserved.