Neural network-based learning of sleep patterns and application-driven interventions

Abstract

As societies advanced, more people suffer from insomnia. In 2014, the US Centers for Disease Control and Prevention declared that sleep disorders should be dealt with as a public health epidemic. However, it is hard to provide adequate treatment for each insomnia sufferer, since various behavioral characteristics influence symptoms of insomnia collectively. This work is composed of three sub-studies, which are clustering insomnia sufferers based on their latent traits extracted from behavior and sleep patterns, predicting sleep efficiency based on the past behavior and sleep information, and designing a mobile intervention to provide a more tailored way of intervention contents to insomnia sufferers, respectively. From the first study, Our method identified 5 new insomnia-activity clusters of participants that conventional methods have not recognized, and significant differences in sleep and behavioral characteristics were shown among groups. From the second study, we predicted the sleep efficiency of individual users with a proposed interpretable LSTM-Attention (LA Block) neural network model. From the third study, we built an application-based cognitive-behavioral treatment for insomnia (CBT-i) intervention named "Sleeps" that aims to alleviate insomnia symptoms in people's daily lives. Our research suggests that a neural network-based computational approach allows health practitioners to devise precise and tailored interventions at the level of data-guided user clusters and predictions on sleep quality (i.e., precision psychiatry), which could be a novel solution to treating insomnia and other mental disorders.