At Mount Sinai, a transformer-based AI model now processes entire nights of sleep data, potentially transforming sleep research and diagnostics by offering unparalleled accuracy and insights.
Researchers at the Icahn School of Medicine at Mount Sinai have developed an advanced AI tool designed to analyze an entire night’s sleep with unprecedented accuracy. Dubbed the Patch Foundational Transformer for Sleep (PFTSleep), this transformative AI model has the potential to revolutionize how sleep disorders are diagnosed and treated.
Published in the journal Sleep, the study is one of the largest of its kind, examining over 1 million hours of sleep data. The PFTSleep model employs a transformer architecture similar to those used in large language models such as ChatGPT, but specifically tailored to process extensive sleep signals.
“This is a step forward in AI-assisted sleep analysis and interpretation,” first author Benjamin Fox, a doctoral candidate at the Icahn School of Medicine at Mount Sinai in the Artificial Intelligence and Emerging Technologies Training Area, said in a news release. “By leveraging AI in this way, we can learn relevant clinical features directly from sleep study signal data and use them for sleep scoring and, in the future, other clinical applications such as detecting sleep apnea or assessing health risks linked to sleep quality.”
Comprehensive Sleep Analysis
Traditional sleep analysis often relies on human experts to manually score short segments of sleep data, or on AI models that can only analyze brief periods.
However, this new approach analyzes complete nights of sleep data, including brain waves, muscle activity, heart rate and breathing patterns, thereby capturing more detailed and nuanced sleep patterns.
The model’s comprehensive analysis promises to standardize sleep research and support clinical tools aimed at identifying sleep disorders and other health risks.
“Our findings suggest that AI could transform how we study and understand sleep,” added co-senior corresponding author Ankit Parekh, an assistant professor of medicine at at the Icahn School of Medicine at Mount Sinai and director of the Sleep and Circadian Analysis Group at Mount Sinai. “Our next goal is to refine the technology for clinical applications, such as identifying sleep-related health risks more efficiently.”
Impact on Clinical Practice
The innovative AI tool is particularly significant in enhancing the efficiency and consistency of sleep studies. While it will not replace clinical expertise, it aims to aid sleep specialists by providing faster and more standardized analyses.
“This AI-driven approach has the potential to revolutionize sleep research,” added co-senior corresponding author Girish N. Nadkarni, the chair of the Windreich Department of Artificial Intelligence and Human Health at Mount Sinai. “By analyzing entire nights of sleep with greater consistency, we can uncover deeper insights into sleep health and its connection to overall well-being.”
Future Directions
Next steps for the researchers include expanding the AI’s capabilities beyond sleep-stage classification to detect specific sleep disorders and to predict various health outcomes. This advancement is vital for creating new diagnostic tools and treatments that could improve patients’ lives.