Medically reviewed by
Dacelin St Martin, MD
Triple board-certified in Sleep Medicine,
Internal Medicine, and Pediatrics.
Artificial intelligence is a big thing at the moment, and it’s not just a fad; it’s a reality of the level of advancement available in the present age.
The history of AI and modern computations dates back to the 1950s when a British logician and mathematician worked on using computers for problem-solving and a ‘thinking’ computer system, giving birth to the term “Turing test.”
As described by a study, artificial intelligence refers to computers’ capacity to execute tasks traditionally deemed unachievable by human intelligence. It includes actions like pattern recognition and decision-making, to mention a few.
The role of AI in healthcare is not to be taken lightly, especially in diagnosing and caring for sleep disorders in the emerging interconnected world we live in.
Impact of AI on Healthcare
The emergence of massive datasets, better algorithms, and faster computers have paved the way for AI/ML (machine learning) applications in various sectors, including healthcare.
For instance, AL/ML applications are proven helpful in identifying early cancer growth (breast and colon cancers), recognizing retinal diseases that could cause blindness, and evaluating and treating certain sleep conditions.
AI’s Role in Sleep Medicine
Experts had earlier predicted that artificial intelligence and machine learning could readily help detect and treat many evasive sleep conditions. With advances in computing technology, this is fast becoming a reality.
An excellent study published in the Journal of Clinical Sleep Medicine demonstrates that machine learning models effectively diagnose Obstructive Sleep Apnea (OSA) by using features from the electrocardiogram, pulse oximetry, and acoustic signals, among other sources. It also revealed that Machine learning (ML) performed well in categorizing OSA and classifying the severity of the patient’s condition.
Furthermore, the study showed that ML accurately predicted surgical treatment and continuous positive airway pressure (CPAP) therapy outcomes.
Currently, sleep physicians offer two types of testing for suspected cases of sleep apnea. These are polysomnograms and home sleep testing. Emerging evidence suggests that both tests can benefit significantly from machine learning, tracking, monitoring, and creating new metrics and data that AI can accomplish.
Dr. Jerry Hu – a triple board-certified sleep expert at the Medical University of South Carolina, alluded that artificial intelligence can make investigative modalities like polysomnography more efficient at detecting OSA.
“Artificial intelligence (AI) has the potential to transform polysomnography (PSG) in many ways, including by greatly improving speed and accuracy, summarizing large amounts of data, recognizing trends, using global data to compare and design new clinical trials, classifying sleep levels and degrees, and developing predictive analytics that previously did not exist,” Dr. Hu said.
Regarding AI’s importance to home sleep testing, Dr. Jerry suggested that AL/ML can help reduce human errors.
“Since patients are responsible for independently setting up and activating their home sleep tests, AI can also enhance data gathering and processing to reduce the frequency of human errors,” he said.
Furthermore, Artificial intelligence (AI) can immediately provide feedback on the placement of home sleep testing (HST) technical and physical components and determine their reliability.
Experts have also suggested in a recent study that AI can help categorize sleep disorder patterns and severity. The study’s findings revealed that AI could classify undiagnosed sleep apnea with an accuracy of up to 83 percent.
Beyond AI’s role in data interpretation, it can also help streamline clinical processes via the automation of steps and introduce better precision in evaluation and treatment.
Studies have shown that using AI for sleep staging can help diagnose Type-1 Narcolepsy with an accuracy higher than any human-dependent scoring system. It demonstrates the potential of AI to automate the scoring of sleep and sleep events while still being able to extract meaningful insights from the data.
AI in Healthcare: Dangers and Fears
A concern with AI use is the perpetuation of existing social bias and subsequent generalization due to the data the AI model is trained with, which can significantly affect the assessment and overall care for underserved demographics and application of stereotypes to the individual from inadequate data or ineffective (wrong methodology) data about their group.
Also, since AI is trained with data of individuals, there are concerns about the safety of indiscriminate access to health records, especially in the absence of clearly defined regulatory functions due to the recent rise in technology and the risk of unintended harmful or pervasive use.
The Future of AI
The future of AI in sleep medicine is full of possibilities. Its use in healthcare will inevitably increase as we push the frontiers of knowledge and technology.
Regarding the future of AI in Obstructive Sleep Apnea (OSA) management, AI can help improve the current diagnostic and treatment modalities and help health professionals personalize treatment for their patients.[4,7]
Dr. Kelvin Postol, president of the American Academy of Dental Sleep Medicine, believes doctors will employ AI to assess patient characteristics and decide which treatments will be most effective or beneficial for each patient.
“Healthcare providers will likely utilize artificial intelligence in the future to maximize efficiencies and reduce superfluous tasks,” he said.
Dr. Postol, however, cautions that AI shouldn’t be used in isolation or replace human clinical acumen. He noted that it is unlikely that AI will replace the expertise of a physician in making a diagnosis and administering treatment.
“Optimal patient care involves humanistic and compassionate care, neither of which can be achieved by artificial intelligence,” he said.
Ultimately, the extent and potential of AI are still grey, and how much it can do cannot be wholesomely assessed at the moment, but one thing is clear as daylight: the horizon is limitless.
- Bandyopadhyay, A., & Goldstein, C. (2023). Clinical applications of artificial intelligence in sleep medicine: a sleep clinician’s perspective. Sleep And Breathing, 27(1), 39–55. https://doi.org/10.1007/s11325-022-02592-4Lovejoy, C. A., Abbas, A. R., & Ratneswaran, D. (2021). An introduction to artificial intelligence in sleep medicine. Journal of thoracic disease, 13(10), 6095–6098. https://doi.org/10.21037/jtd-21-1569
- Bazoukis, G., Bollepalli, S. C., Chung, C. T., Li, X., Tse, G., Bartley, B. L., Batool-Anwar, S., Quan, S. F., & Armoundas, A. A. (2023). Application of artificial intelligence in the diagnosis of sleep apnea. Journal of Clinical Sleep Medicine: JCSM: Official Publication of the American Academy of Sleep Medicine. https://doi.org/10.5664/jcsm.10532
- Goldstein, C. A., Berry, R. B., Kent, D. T., Kristo, D. A., Seixas, A. A., Redline, S., & Westover, M. B. (2020). Artificial intelligence in sleep medicine: background and implications for clinicians. Journal of clinical sleep medicine: JCSM: official publication of the American Academy of Sleep Medicine, 16(4), 609–618. https://doi.org/10.5664/jcsm.8388
- Molnár, V., Kunos, L., Tamás, L., & Lakner, Z. (2023). Evaluation of the applicability of artificial intelligence for the prediction of obstructive sleep apnoea. Applied Sciences (Basel, Switzerland), 13(7), 4231. https://doi.org/10.3390/app13074231
- Stephansen, J. B., Olesen, A. N., Olsen, M., Ambati, A., Leary, E. B., Moore, H. E., Carrillo, O., Lin, L., Han, F., Yan, H., Sun, Y. L., Dauvilliers, Y., Scholz, S., Barateau, L., Hogl, B., Stefani, A., Hong, S. C., Kim, T. W., Pizza, F., … Mignot, E. (2018). Neural network analysis of sleep stages enables efficient diagnosis of narcolepsy. Nature Communications, 9(1), 5229. https://doi.org/10.1038/s41467-018-07229-3
- Brennan, H. L., & Kirby, S. D. (2023). The role of artificial intelligence in the treatment of obstructive sleep apnea. Le Journal d’oto-Rhino-Laryngologie et de Chirurgie Cervico-Faciale [Journal of Otolaryngology – Head & Neck Surgery], 52(1), 7. https://doi.org/10.1186/s40463-023-00621-0.