Publication: AI-Driven Sleep Staging Using Instantaneous Heart Rate and Accelerometry: Insights from an Apple Watch Study

Authors: Tzu-An Song, Yubo Zhang, Ziyuan Zhou, Luke Hou, Masoud Malekzadeh, Aida Behzad, Joyita Dutta Abstract Polysomnography, the gold standard for sleep evaluations, involves complex setup and data acquisition protocols and requires manual scoring of sleep data. Smartwatches and other multi-sensor consumer wearable devices with automated sleep staging capabilities offer a promising and scalable alternative for routine and long-term sleep evaluations in individuals. We conducted a multi-night study using a smartwatch for sleep assessment and created an AI-driven automated sleep staging framework based on instantaneous heart rate (IHR) and accelerometry data using sleep stage labels based on electroencephalography (EEG) as the reference. 47 healthy adults were recruited to record their sleep for up to seven consecutive nights using an Apple Watch Series 6 and a Dreem 2 Headband. Our sleep staging framework relies on a…

Continue ReadingPublication: AI-Driven Sleep Staging Using Instantaneous Heart Rate and Accelerometry: Insights from an Apple Watch Study

New Product Launch: BIDSleep iPhone and Apple Watch App

BIDSleep helps you access wellness data from your Apple Watch, including heart rate, motion, blood oxygen, and sleep stages, all securely stored on your device. App Purpose: BIDSleep is a wellness app that helps users collect heart rate, blood oxygen (SpO₂), motion, and optional sleep stage data during rest or overnight sessions. It is designed solely for personal data logging and self-monitoring purposes. Source: https://apps.apple.com/us/app/bidsleep/id6747012248?platform=iphone

Continue ReadingNew Product Launch: BIDSleep iPhone and Apple Watch App

Open Source AI-model Released: SLAMSS-IFS

To the study team's knowledge, this is the first open-source four-class sleep staging model developed from a multi-night Apple Watch sleepstudy. SLAMSS-IFS, an advanced version of our previous SLAMSS model, for four-class sleep staging using IHR and accelerometry signals fromthese wearable devices. Key innovations in the model, including an intra-epoch learning LSTM, frequency information incorporation, andskip connections, contribute to substantial performance improvements over other SLAMSS variants and other state-of-the-art models. Ourresults show that SLAMSS-IFS outperforms competing models in overall accuracy, sensitivity, specificity, precision, weighted F1 score, weighted MCC, and most clinical sleep metrics. SLAMSS-IFS: The SLAMSS-IFS model builds on the original SLAMSS model with three additional components: an intraepoch learning sequence-to-sequence long short-term memory (LSTM (“I”), a frequency variable (“F”), and a skip connection (“S”). The intra-epoch learning LSTM processes temporal dependencies within individual epochs.…

Continue ReadingOpen Source AI-model Released: SLAMSS-IFS

Grant Funding: R01 AG082354

Title: Genomics-guided sleep biomarker discovery for early Alzheimer's disease: A wearables study This R01 builds upon the technology and algorithms for sleep-based metrics developed in the MassAITC pilot project, utilizing the same EEG device. It shifts the study from a general AD-risk factor population to a genetic AD-risk factor population. Public Health Relevance Statement: Sleep-based metrics are promising as potential noninvasive biomarkers for the early identification of individuals at risk of Alzheimer’s disease. In this R01 Research Project Grant, we will collect and analyze sleep and activity data from three wearable devices for sleep, heart rate, and activity monitoring along with blood biomarkers for Alzheimer’s disease from elderly Mass General Brigham Biobank participants with elevated genetic susceptibility to Alzheimer’s disease. We will also develop artificial intelligence tools for digital phenotyping and the discovery of novel…

Continue ReadingGrant Funding: R01 AG082354