Doctors use AI to detect Sleep Apnea

Two recent advances in health technology may make diagnosing sleep apnea easier in the near future. One involves applying the power of artificial intelligence (AI) to X-rays of the head and neck to identify if a person is likely to have sleep apnea The other is a swallowable, pill-sized capsule that can monitor breathing and heart rate from inside the body.  

Finding new and better ways to identify and diagnose people with sleep apnea is significant because over 80% of those with moderate to severe sleep apnea are undiagnosed. Without treatment, these people face a greater risk of anxiety, depression, heart attack, stroke, car crashes, and other problems. 

Many people with sleep apnea go undiagnosed because of a lack of adequate screening tools. The existing tools are unreliable, and many doctors don’t use them.

Dr. Han-Gil Jeong led a team of researchers from Seoul National University in testing out a potential new sleep apnea screening tool earlier this year. After using AI technology to analyze X-rays of people’s heads and necks, this tool could predict whether or not a person had sleep apnea with a high level of accuracy.

Obstructive sleep apnea occurs because the upper airways become narrowed or blocked while a person sleeps. By learning from past X-rays of people with obstructive sleep apnea, the team’s AI model could find slight tongue, mouth, or throat variations that might lead to a blocked airway. These features are so small that experts can’t identify them by simply looking at the X-rays with the human eye. 

A lack of adequate screening tools isn’t the only reason people with sleep apnea go undiagnosed. Even people identified as likely having sleep apnea may not be screened because testing can be complex and inconvenient. 

Sleep apnea is currently diagnosed with an overnight sleep study in a lab or at home. Both require a person to stay hooked up to various monitoring devices while they sleep, which may make testing difficult, especially for people with certain heart or lung conditions.

Researchers from the Massachusetts Institute of Technology and West Virginia University recently tried an ingestible capsule that collects heart and breathing data in humans for the first time. 

Participants in the research study swallowed the capsules and then spent the night undergoing traditional sleep studies. This approach allowed researchers to compare the capsules’ data to those from the usual monitoring devices. The capsules accurately collected breathing and heart rate data that closely mirrored that collected by the sleep study devices. This demonstrates the capsule’s potential for eventually replacing some monitoring devices now used in sleep apnea testing.

Got a hot tip? Pitch us your story idea, share your expertise with, or let us know about your sleep experiences right here.

Learn more about our Editorial Team

5 Sources

  1. Jeong, H., Kim, T., Hong, J., Kim, H. J., Yun, S., Kim, S., Yoo, J., Lee, J. Y., Thomas, R. J., & Yun, C. (2023). Automated deep neural network analysis of lateral cephalogram data can aid in detecting obstructive sleep apnea. Journal of Clinical Sleep Medicine, 19(2), 327–337.
  2. Sleep Review. (2023, November 16). AI-Powered X-Ray analysis can diagnose sleep apnea. Sleep Review.
  3. Traverso, G., Finomore, V., Mahoney, J. J., Kupec, J., Stansbury, R., Bacher, D., Pless, B., Schuetz, S. M., Hayward, A., Langer, R., & Rezai, A. R. (2023). First-in-human trial of an ingestible vitals-monitoring pill. Device, 1(5), 100125.
  4. Kline, L. R. (2023, October 5). Clinical presentation and diagnosis of obstructive sleep apnea in adults. In N. Collop (Ed.). UpToDate.
  5. Collop, N. (2023, June 7). Home sleep apnea testing for obstructive sleep apnea in adults. In S. M. Harding (Ed.). UpToDate.

Learn More About Sleep News