The Potential and Limitations of AI Algorithms in Sleep Care: Anuja Bandyopadhyay, MD
The assistant professor of clinical pediatrics at Indiana University School of Medicine, and the chair of the AASM’s Artificial Intelligence in Sleep Medicine Committee, spoke about the ability of AI algorithms as tools in clinical care. [WATCH TIME: 3 minutes]
WATCH TIME: 3 minutes
“An AI algorithm, for the most part, is as good as the dataset it gets trained on. For instance, if I am training it with data from a 4-year-old [population] who have a certain percentage of sleep and are normally healthy, after training my machine on that population if I then take that algorithm and use it on say, a 60-year-old [population] with multiple morbidities, it may not work as well.”
In sleep medicine, there has been a recent push from the American Academy of Sleep Medicine (AASM) to incorporate artificial intelligence (AI) algorithms and tools into clinical care. As of now, these are limited somewhat in their capacity but still helpful, being used mostly for sleep staging and monitoring, which are currently very labor intensive. The use of these AI tools could, as a result, reduce the time needed to collect rich sleep stage data.
AASM has established a 2-year pilot program, call the AI/Autoscoring Certification Program, which will independently verify the performance of these auto-scoring systems.1 Currently, the focus is solely on polysomnograms for sleep stage scoring, with the software being assessed needing to demonstrate accuracy that is equivalent to or better than manual scoring to be certified, with applications to the program accepted starting in late 2022 or early 2023.
NeurologyLive® spoke with Anuja Bandyopadhyay, MD, assistant professor of clinical pediatrics, Indiana University School of Medicine, and the chair of the AASM’s Artificial Intelligence in Sleep Medicine Committee to learn more about what these AI algorithms could do for those in the sleep disorder field. She spoke to the potential of these programs, as well as the importance of understanding their limitations and the importance of the datasets they’re trained on.
Per the AASM, companies using auto-score technology that would like to learn more about the AI/Autoscoring Pilot Certification Program, can contact the AASM at AIGSP@aasm.org.
