Related Experiment Video
Updated: Jun 23, 2026
06:49
Evaluation of a Smartphone-based Human Activity Recognition System in a Daily Living Environment
Published on: December 11, 2015
Accuracy of Smartphone-Mediated Snore Detection in a Simulated Real-World Setting: Algorithm Development and
Jeffrey Brown1, Zachary Mitchell1, Yu Albert Jiang1
1Bodymatter, Inc, 4343 Von Karman Ave, Suite 150J, Newport Beach, CA, 92660, United States, 1 877-870-0649.
JMIR Formative Research
|March 28, 2025
Summary
The SleepWatch app accurately detects snoring using machine learning, achieving 95.6% overall accuracy in simulated tests. This tool may help identify individuals at risk for sleep apnea.
Area of Science:
- Sleep science
- Digital health technology
- Machine learning applications
Background:
- High-quality sleep is crucial for physical and mental health.
- Poor sleep is linked to various health issues, including cardiometabolic diseases and increased mortality.
- Snoring disrupts sleep and is associated with conditions like obstructive sleep apnea.
Purpose of the Study:
- To evaluate the accuracy of the SleepWatch smartphone app's snore detection algorithm.
- To assess the algorithm's performance in a simulated real-world environment.
Main Methods:
- The SleepWatch algorithm was tested using 36 simulated snoring audio files (30-600 snores/hour) and 9 nonsnoring files.
- Performance metrics included sensitivity, specificity, accuracy, positive predictive value, and negative predictive value.
- Bland-Altman plots and Spearman correlation were used for statistical analysis.
Main Results:
- The algorithm achieved an average accuracy of 95.2% for snoring tests and 97.1% specificity for nonsnoring sounds.
- Overall aggregated accuracy across all tests was 95.6%.
- Strong positive correlation (rs=0.974; P<.001) was found between detected and actual snore rates.
Conclusions:
- The SleepWatch snore detection algorithm demonstrates high accuracy and reliability.
- It performs comparably to other snore detection applications.
- The app shows potential for identifying individuals at risk for sleep-disordered breathing, such as obstructive sleep apnea, based on snoring index.
