Key result
An explainable AI framework using smartwatch lifelog data achieved an 80% accuracy rate in predicting sleep efficiency and identified lifestyle adjustments to improve sleep quality.
Why the study?
Wearable devices primarily monitor sleep patterns without offering strategies to improve sleep efficiency.
Observational
An interpretable AI framework using smartwatch data can predict sleep efficiency with 80% accuracy and suggest actionable lifestyle changes to improve sleep quality.
This is the authors' abstract. We don't add key points for this paper.
BACKGROUND Sleep plays a crucial role in the immune system, memory, and emotional control, which in turn affects overall quality of life. Although several studies have used wearable devices to measure sleep efficiency, these tools primarily monitor sleep patterns without offering strategies to improve sleep efficiency. OBJECTIVE In response, our study introduced a framework that not only measures sleep efficiency using wearable technology but also suggests lifestyle changes to enhance sleep quality, using explainable artificial intelligence. METHODS Lifelog data was measured using a smartwatch and only features that can be controlled by the user, such as the number of steps, were extracted from the lifelog data. A light gradient-boosted machine was utilized as a classifier to predict whether sleep efficiency was low or high. Using Shapley additive explanation, important features related to sleep efficiency, such as walking speed were identified and lifestyles were recommended to improve sleep efficiency. We also investigated the relationship between sleep efficiency and sleep stages, which were excluded from the features extracted as uncontrolled factors from wearable devices. RESULTS Results show that, following a 10-fold leave-one-subject-out cross-validation evaluation, an accuracy rate of 80% was achieved in measuring sleep efficiency. In a scenario where lifestyle was adjusted according to the feature importance derived from the model output, an improvement from bad to good sleep efficiency was observed. In addition, the sleep stage variables were found to be statistically higher for the cases exhibiting good sleep efficiency. CONCLUSIONS This framework provides insights into developing technologies that suggest lifestyle changes to enhance sleep efficiency, potentially leading to better health outcomes through improved sleep quality.
Answers are grounded in this paper and cite their sources.
Supports wearable AI sleep monitoring in research; leaves open clinical validation and outcome impact before practice adoption.
No takes yet. Share an insight, caveat, or question.
Jeong et al. (2024) conducted an observational in Sleep efficiency. Explainable AI framework using smartwatch lifelog data was evaluated on Accuracy in predicting whether sleep efficiency was low or high. An explainable AI framework using smartwatch lifelog data achieved an 80% accuracy rate in predicting sleep efficiency and identified lifestyle adjustments to improve sleep quality.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: