Evaluating depression with multimodal wristband-type wearable device: screening and assessing patient severity utilizing machine-learning

Abstract

We aimed to develop a machine learning algorithm to screen for depression and assess severity based on data from wearable devices. Methods: We used a wearable device that calculates steps, energy expenditure, body movement, sleep time, heart rate, skin temperature, and ultraviolet light exposure. Depressed patients and healthy volunteers wore the device continuously for the study period. The modalities were compared hourly between patients and healthy volunteers. XGBoost was used to build machine learning models and 10-fold cross-validation was applied for the validation. Results: Forty-five depressed patients and 41 healthy controls participated, creating a combined 5,250 days' worth of data. Heart rate, steps, and sleep were significantly different between patients and healthy volunteers in some comparisons. Similar differences were also observed longitudinally when patients' symptoms improved. Based on seven days' data, the model identified symptomatic patients with 0.76 accuracy and predicted Hamilton Depression Rating Scale-17 scores with a 0.61 correlation coefficient. Skin temperature, sleep time-related features, and the correlation of those modalities were the most significant features in machine learning. Limitations: The small number of subjects who participated in this study may have weakened the statistical significance of the study. There are differences in the demographic data among groups although we performed a correction for multiple comparisons. Validation in independent datasets was not performed, although 10-fold cross validation with the internal data was conducted. Conclusion:The results indicated that utilizing wearable devices and machine learning may be useful in identifying depression as well as assessing severity.

Current Advances in Wearable Devices and Their Sensors in Patients With Depression

“…Among the collected studies on wearable devices for patients with depression, two-thirds used actigraph units, while the rest used other devices such as a novel wearable device with three accelerometers (18); medically used wearable devices, such as a Parkinson's KinetiGraph (PKG) (23) and the E4 wristband (Empatica, Boston, MA, USA) (25); and commercial wearable devices not originally intended for medical use, such as the Fitbit Flex TM (Fitbit, Inc., San Francisco, CA, USA) (19), Apple watch (Apple Inc., Cupertino, CA, USA) (20), and Silmee TM W20 wristband (TDK Corporation, Tokyo, Japan) (22). Among the reviewed studies, no wearable devices were used for treatment.…”

“…It contains a small touchscreen, which was used by Cormack et al (20) to offer tasks like cognitive and mood assessments to study participants. Finally, the Silmee TM W20 wristband was used previously (22) for unintended medical use and is equipped with a three-axis acceleration sensor, pulse sensor, ultraviolet sensor, and temperature sensor.…”

Analysis of Data from Wearable Sensors for Sleep Quality Estimation and Prediction Using Deep Learning

“…The attention-based deep learning model shows improvement of performance. Tazawa et al [49] develop a machine learning algorithm to screen for depression and assess its severity based on data from wearable devices. The use of machine learning and deep learning techniques for automation of sleep quality prediction has also been studied by some researchers.…”

Unobtrusive Sensing Technology for Quantifying Stress and Well-Being Using Pulse, Speech, Body Motion, and Electrodermal Data in a Workplace Setting: Study Concept and Design

“…Not only using single modality but combining multimodal data and with machine learning approach it may be more realistic to screen depression or to predict severity of depression. Utilizing wrist band-type wearable device that record three-axis acceleration, heart rate, body temperature, and ultraviolet light exposure, we have reported that it was possible to identify patients with depression with an accuracy of 0.76, and to predict depression severity with a 0.61 correlation coefficient with Hamilton Depression Rating Scale score ( 15 , 18 ).…”