Smartphone knowledge helps predict schizophrenia relapses

Passive data from smartphones – including movement, ambient noise and sleep patterns – can help predict episodes of schizophrenic relapse, according to a new study from Cornell Tech.

Two papers from the laboratory of Tanzeem Choudhury, Professor of Integrated Health and Technology at Cornell Tech, examined how smartphone data can predict patients’ self-assessment of their condition and changes in behavioral patterns over the 30 days that lead to illness lead relapse.

Predicting schizophrenic relapses early – potentially dangerous episodes that could include hallucinations, fear of harm, depression, or withdrawal – could prevent hospital stays and provide doctors and patients with valuable information that could improve and personalize their care.

“The goal of this work was to predict digital indicators that are early warning signs of relapse, but these symptoms or changes can vary greatly from person to person,” said Dan Adler, PhD student at Cornell Tech and lead author of “Prediction of.” Early Warning Signs of Psychotic Relapse from Passive Acquisition Data: An Approach Using Neural Encoder-Decoder Networks, “published in the Journal of Medical Internet Research, mHealth and uHealth.

“We tried to develop an approach that would allow us to tell a clinician: Not only is this participant experiencing unusual behavior, these are the specific things that are different in that particular patient,” Adler said. “If we can predict when a person’s symptoms will change before relapse, we can get them treated early and possibly prevent an inpatient visit.”

The researchers collected smartphone data from 60 participants over a year, 18 of whom had relapsed during this time. They used encoder-decoder neural networks – a type of machine learning that is good at learning complex functions amid very irregular data – to detect behavioral patterns such as sleep, number of missed calls, and the duration and frequency of calls.

The method found a mean increase in behavioral abnormalities of 108% in the 30 days prior to relapses compared to behavior on days of relative health.

The paper used data collected in collaboration with the University of Washington, Dartmouth College, and the Northwell Health System. Based on the same dataset, another paper – “Using Behavioral Rhythms and Multitasking Learning to Predict Fine-Grained Symptoms of Schizophrenia” published in Scientific Reports – used machine learning to better understand and predict symptoms from changes in behavioral rhythms passively from smart devices recognized.

“We wanted to provide some actionable steps or clinically interpretable characteristics so we could either tell the patient to take action or tell the clinician to intervene early,” said Vincent Tseng, PhD student at Cornell Tech, Co- First author of the Scientific Reports paper.

In digital phenotyping, smartphones can play a role in assessing severe mental illness

More information:
Daniel A. Adler et al., Predicting Early Warning Signs of Psychotic Relapse Using Passive Sensor Data: An Approach Using Encoder-Decoder Neural Networks, JMIR mHealth and uHealth (2020). DOI: 10.2196 / 19962 Provided by Cornell University

Quote: Smartphone data helps predict schizophrenia relapses (2020, October 13th), retrieved from on October 13, 2020

This document is subject to copyright. Except for fair trade for the purpose of private study or research, no part may be reproduced without written permission. The content is provided for informational purposes only.

Comments are closed.