Smartwatch-Based AI Model for OSA Prediction (SWOSA)

Smartwatch-Based Artificial Intelligence Model for Obstructive Sleep Apnea Prediction

This study aims to develop an artificial intelligence (AI) model for more accurately diagnosing obstructive sleep apnea (OSA) by collecting blood oxygen saturation and other health information during sleep using a smartwatch.

OSA is common but often underdiagnosed, and the gold-standard diagnostic test, polysomnography, is costly and time-consuming. Smartwatches can provide a variety of health data, such as sleep patterns, blood oxygen saturation, and heart rate, which can help detect key symptoms and signs of OSA.

By developing an AI model that uses smartwatch data to screen for OSA, this study seeks to offer a cost-effective and accessible diagnostic method, ultimately contributing to the early detection and improved treatment rates of OSA.

Study Overview

• Status: Recruiting
• Conditions: Obstructive Sleep Apnea of Adult; Screening; Smart Watch
• Intervention / Treatment: Device: Galaxy Watch 4, Samsung Electronics Co., Ltd., South Korea

• Study Type: Observational
• Enrollment (Estimated): 147

Contacts and Locations

• Study Contact: Name: Jaeyoung Cho, M.D., Ph.D.; Phone Number: +82-2-2072-2503; Email: apricot6@snu.ac.kr

Study Locations

• Korea, Republic of
  • Seoul, Korea, Republic of, 03080
    • Recruiting
    • Seoul National University Hospital
    • Contact: Jaeyoung Cho, MD; Phone Number: 82-2-2072-2503; Email: apricot6@snu.ac.kr

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: No

• Sampling Method: Non-Probability Sample

Study Population

Men and women aged 22 to 85 years who visited Seoul National University Hospital with suspected sleep apnea due to symptoms such as snoring, apnea, or excessive daytime sleepiness.

Description

Inclusion Criteria:

• Men and women aged 22 to 85 years who visited Seoul National University Hospital with suspected sleep apnea due to symptoms such as snoring, apnea, or excessive daytime sleepiness.

Exclusion Criteria:

• Patients previously diagnosed with sleep apnea who are currently undergoing treatment (e.g., positive airway pressure [PAP] therapy, mechanical ventilation, oral appliances, or surgery).
• Patients with neuromuscular diseases or a history of chronic opioid medication use.
• Patients with severe insomnia that is not controlled by medication.
• Patients receiving supplemental oxygen therapy due to underlying conditions such as heart failure, chronic obstructive pulmonary disease, interstitial lung disease, hypoventilation syndrome, or stroke, or whose baseline oxygen saturation is less than 90%.
• Patients with implanted cardiac pacemakers, defibrillators, or other electronic devices.
• Patients inexperienced in using smartphones, apps, or smartwatches.
• Pregnant women.
• Patients unable or unwilling to provide written informed consent.

Study Plan

How is the study designed?

Number of groups / cohorts

1

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
Smart Watch Group; Men and women aged 22 to 85 years who visited Seoul National University Hospital with suspected sleep apnea due to symptoms such as snoring, apnea, or excessive daytime.	Device: Galaxy Watch 4, Samsung Electronics Co., Ltd., South Korea; Use of the Galaxy Watch 4 during sleep for approximately two weeks prior to the polysomnography test, including the night of the test.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Predictive Accuracy of the AI Model for Moderate-to-Severe Obstructive Sleep Apnea	Evaluation of how well the AI model, developed using clinical data and smartwatch-recorded information including nocturnal oxygen saturation, predicts moderate-to-severe obstructive sleep apnea (defined as apnea-hypopnea index ≥15/hour) diagnosed by polysomnography.	Up to 2 weeks prior to the polysomnography test.

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Predictive Accuracy of the Galaxy Watch Sleep Apnea Feature (SAF)	Assessment of the accuracy of the Galaxy Watch's built-in sleep apnea feature (SAF) in predicting moderate-to-severe obstructive sleep apnea diagnosed by polysomnography.	Up to 2 weeks prior to the polysomnography test.
Comparison of AI Model and Galaxy Watch Sleep Apnea Feature (SAF) Performance	Comparison of the predictive performance between the AI model developed in this study and the Galaxy Watch's built-in sleep apnea feature (SAF) for detecting moderate-to-severe obstructive sleep apnea.	Up to 2 weeks prior to the polysomnography test.
Comparison of AI Model and STOP-Bang Questionnaire Performance	Comparison of the predictive performance between the AI model developed in this study and the STOP-Bang questionnaire for detecting moderate-to-severe obstructive sleep apnea.	Up to 2 weeks prior to the polysomnography test.

Collaborators and Investigators

• Sponsor: Seoul National University Hospital
• Investigators: Principal Investigator: Jaeyoung Cho, M.D., Ph.D., Seoul National University Hospital

Study record dates

Study Major Dates

• Study Start (Actual): February 3, 2025
• Primary Completion (Estimated): December 31, 2026
• Study Completion (Estimated): December 31, 2026

Study Registration Dates

• First Submitted: January 19, 2025
• First Submitted That Met QC Criteria: January 19, 2025
• First Posted (Actual): January 24, 2025

Study Record Updates

• Last Update Posted (Actual): May 15, 2025
• Last Update Submitted That Met QC Criteria: May 12, 2025
• Last Verified: May 1, 2025

More Information

Terms related to this study

• Additional Relevant MeSH Terms: Nervous System Diseases; Respiratory Tract Diseases; Respiration Disorders; Sleep Wake Disorders; Apnea; Sleep Disorders, Intrinsic; Dyssomnias; Sleep Apnea Syndromes; Sleep Apnea, Obstructive
• Other Study ID Numbers: 24111291590

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: YES
• IPD Plan Description: De-identified individual participant data that support the findings of this study
• IPD Sharing Supporting Information Type: STUDY_PROTOCOL

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No
• product manufactured in and exported from the U.S.: No