Belun Ring for Sleep Apnea

Recruiting at 1 trial location

Overseen By Tiffany Tsai

Age: 18+

Sex: Any

Trial Phase: Academic

Sponsor: Belun Technology Company Limited

Disqualifiers: Atrial fibrillation, Pacemaker, Recent surgery, others

No Placebo Group All trial participants will receive the active study treatment (no placebo)

Approved in 1 Jurisdiction This treatment is already approved in other countries

Compensation: Varies

Number of Visits: Unknown

Duration: 2 years

79 Participants Needed

Trial Summary

Will I have to stop taking my current medications?

The trial information does not specify whether you need to stop taking your current medications. It's best to discuss this with the trial coordinators or your doctor.

What data supports the effectiveness of the Belun Ring treatment for sleep apnea?

Research shows that the Belun Ring, using deep learning algorithms, can effectively detect obstructive sleep apnea (OSA) and classify sleep stages. It has been validated against standard sleep studies, demonstrating its ability to assess OSA severity and provide accurate sleep data at home. ^{1 2 3 4 5}

How is the Belun Ring treatment different from other sleep apnea treatments?

The Belun Ring is a unique wearable device that uses advanced deep learning algorithms to detect and categorize sleep apnea severity and classify sleep stages, allowing for home-based testing instead of traditional sleep lab studies. ^{1 2 3 6 7}

What is the purpose of this trial?

Hypothesis: BR's Gen3 DL algorithms, combined with its subxiphoid body sensor, can accurately diagnose OSA, categorize its severity, identify REM OSA and supine OSA, and detect central sleep apnea (CSA). Primary Objective: To rigorously evaluate the overall performance of the BR with Gen3 DL Algorithms and Subxiphoid Body Sensor in assessing SDB in individuals referred to the sleep labs with clinical suspicion of sleep apnea and a STOP-Bang score \> 3, by comparing to the attended in-lab PSG, the gold standard. Secondary Objectives: To determine the accuracy of BR sleep stage parameters using the Gen3 DL algorithms by comparing to the in-lab PSG; To assess the accuracy of the BR arrhythmia detection algorithm; To assess the impact of CPAP on HRV (both time- and frequency-domain), delta HR, hypoxic burden, and PWADI during split night studies; To assess if any of the baseline HRV parameters (both time- and frequency-domain), delta heart rate (referred to as Delta HR), hypoxic burden, and pulse wave amplitude drop index (PWADI) or the change of these parameters may predict CPAP compliance; To evaluate the minimum duration of quality data necessary for BR to achieve OSA diagnosis; To examine the performance of OSA screening tools using OSA predictive AI models formulated by National Taiwan University Hospital (NTUH) and Northeast Ohio Medical University (NEOMED).

Research Team

Ambrose A. Chiang, MD

Principal Investigator

University Hospitals Cleveland Medical Center

Susheel P. Patil, MD

Principal Investigator

University Hospitals Cleveland Medical Center

Kingman P. Strohl, MD

Principal Investigator

University Hospitals Cleveland Medical Center

Eligibility Criteria

This trial is for individuals suspected of having sleep apnea, indicated by a clinical assessment and a STOP-Bang score of 3 or higher. Participants must provide informed consent to join the study.

Inclusion Criteria

Provision of signed informed consent form

Clinically assessed and suspicious for OSA with a STOP-Bang score ≥ 3

Exclusion Criteria

Unstable cardiopulmonary status on the night of the study judged to be unsafe for sleep study by the sleep tech and/or the on-call sleep physician

I have not been hospitalized or had surgery in the last 30 days.

If a participant did not sleep for at least 4 hours of technically valid sleep based on the Belun Ring method for diagnostic assessments, or a minimum of 3 hours of technically valid sleep during the diagnostic phase of a split-night study, the patient will be excluded from statistical analysis

Timeline

Screening

Participants are screened for eligibility to participate in the trial

2-4 weeks

Treatment

Participants undergo evaluation of the BR Gen3 DL algorithms and Subxiphoid Body Sensor for diagnosing sleep disordered breathing, compared to in-lab PSG

2 years

Follow-up

Participants are monitored for safety and effectiveness after treatment

4 weeks

Data Analysis

Analysis of biomarker dynamics and performance metrics of the BR device, including sensitivity and specificity for arrhythmia detection and sleep stage classification

3 years

Treatment Details

Interventions

• Belun Ring Gen3 Deep Learning Algorithms

Trial Overview The Belun Ring with Gen3 Deep Learning Algorithms and Subxiphoid Body Sensor is being tested against in-lab polysomnography (PSG) to diagnose sleep apnea, determine its severity, identify specific types of sleep apnea, and detect arrhythmias.

Participant Groups

Group I: UH-Participant Experimental Treatment 2 Interventions

Potential participants will be identified from patients scheduled for in-lab PSG at the two medical centers of University Hospitals

Belun Ring Gen3 Deep Learning Algorithms is already approved in United States for the following indications:

Approved in United States as Belun Sleep System for:

• Diagnosis of obstructive sleep apnea (OSA)

Who Is Running the Clinical Trial?

Belun Technology Company Limited

Lead Sponsor

Trials

8

Recruited

440+

University Hospitals Cleveland Medical Center

Collaborator

Trials

348

Recruited

394,000+

Findings from Research

The Belun Ring, using advanced deep learning algorithms, accurately detected obstructive sleep apnea (OSA) with an overall accuracy of 0.91 for severe cases (AHI ≥ 30) and demonstrated good performance in categorizing OSA severity across different levels.

In addition to OSA detection, the Belun Ring effectively classified sleep stages, achieving an accuracy of 0.90 for REM sleep, indicating its potential as a reliable tool for both diagnosing sleep apnea and monitoring sleep quality.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Belun Ring (Belun Sleep System BLS-100): Deep learning-facilitated wearable enables obstructive sleep apnea detection, apnea severity categorization, and sleep stage classification in patients suspected of obstructive sleep apnea. Strumpf, Z., Gu, W., Tsai, CW., et al.[2023]

The Belun Sleep Platform (BSP), a wearable ring device, shows strong correlation with traditional polysomnography (PSG) in detecting obstructive sleep apnea (OSA), with a correlation coefficient of r = 0.888, indicating it can effectively assess OSA in a home setting.

BSP demonstrated high accuracy (0.808) and sensitivity (0.931) for identifying moderate to severe OSA, even in patients with comorbidities or on heart rate-affecting medications, suggesting it is a reliable tool for clinical use.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Detection of obstructive sleep apnea using Belun Sleep Platform wearable with neural network-based algorithm and its combined use with STOP-Bang questionnaire. Yeh, E., Wong, E., Tsai, CW., et al.[2021]

The Belun Ring Platform demonstrated strong correlation with traditional sleep studies, showing a high accuracy in predicting the apnea-hypopnea index (AHI) and total sleep time in a study of 50 adults without significant health issues.

The device showed good sensitivity (0.85) and specificity (0.87) for identifying moderate to severe sleep apnea (AHI ≥ 15 events/h), although it tended to overestimate AHI in lower ranges and underestimate in higher ranges, indicating some limitations in its accuracy.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Belun Ring Platform: a novel home sleep apnea testing system for assessment of obstructive sleep apnea. Gu, W., Leung, L., Kwok, KC., et al.[2021]

The Somnofy sleep assistant demonstrated high sensitivity (0.97) and reasonable specificity (0.72) for detecting sleep compared to polysomnography (PSG), indicating it can effectively classify sleep stages in healthy individuals.

While Somnofy showed promising results in sleep stage differentiation, it had some limitations, such as underestimating total wake time and N1/N2 sleep, suggesting further research is needed to validate its performance across different populations, including children and those with sleep disorders.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Validation of sleep stage classification using non-contact radar technology and machine learning (Somnofy®). Toften, S., Pallesen, S., Hrozanova, M., et al.[2021]

The study developed four advanced neural network models to predict obstructive sleep apnea (OSA) events 30 seconds in advance using raw respiratory data from over 46,000 examples across 1,507 subjects, achieving an accuracy of 81%.

The best-performing models, 1D-CNN-LSTM and 2D-CNN-LSTM, demonstrated over 83% accuracy, sensitivity, and specificity, indicating that these models can effectively predict OSA events and could be used in devices to regulate breathing during sleep, even with low-frequency data sampling.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

Sleep Apnea Prediction Using Deep Learning. Wang, E., Koprinska, I., Jeffries, B.[2023]

In a study of 129 participants with hypertension and high cardiovascular risk, the Belun Ring demonstrated a strong correlation with traditional polysomnography for diagnosing obstructive sleep apnea (OSA), showing an intraclass correlation coefficient of 0.882.

The Belun Ring achieved high diagnostic accuracy for OSA, with a sensitivity of 95.7% and an overall accuracy of 87.4%, indicating it is a reliable alternative to polysomnography for screening at-risk individuals.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

The Belun sleep platform to diagnose obstructive sleep apnea in patients with hypertension and high cardiovascular risk. Ou, YH., Ong, J., Thant, AT., et al.[2023]

A new portable sleep monitoring model using a bidirectional long and short-term memory network (BI-LSTM) was developed to detect sleep apnea syndrome (SAS) based on single-channel EEG signals from 42 subjects.

The model demonstrated high performance with an accuracy of 92.73% and precision of 84.21%, indicating its potential effectiveness for cost-effective and accessible SAS monitoring.

NCBI (Pubmed)

pubmed.ncbi.nlm.nih.gov

BI - Directional long short-term memory for automatic detection of sleep apnea events based on single channel EEG signal. Wang, Y., Xiao, Z., Fang, S., et al.[2022]

References

1. United States pubmed.ncbi.nlm.nih.gov

Belun Ring (Belun Sleep System BLS-100): Deep learning-facilitated wearable enables obstructive sleep apnea detection, apnea severity categorization, and sleep stage classification in patients suspected of obstructive sleep apnea. [2023]

2. United States pubmed.ncbi.nlm.nih.gov

Detection of obstructive sleep apnea using Belun Sleep Platform wearable with neural network-based algorithm and its combined use with STOP-Bang questionnaire. [2021]

3. United States pubmed.ncbi.nlm.nih.gov

Belun Ring Platform: a novel home sleep apnea testing system for assessment of obstructive sleep apnea. [2021]

4. Netherlands pubmed.ncbi.nlm.nih.gov

Validation of sleep stage classification using non-contact radar technology and machine learning (Somnofy®). [2021]

5. United States pubmed.ncbi.nlm.nih.gov

Sleep Apnea Prediction Using Deep Learning. [2023]

6. Netherlands pubmed.ncbi.nlm.nih.gov

The Belun sleep platform to diagnose obstructive sleep apnea in patients with hypertension and high cardiovascular risk. [2023]

7. United States pubmed.ncbi.nlm.nih.gov

BI - Directional long short-term memory for automatic detection of sleep apnea events based on single channel EEG signal. [2022]