To access the session recording, navigate to the content tab and click the view video button. 

Session Schedule

Find a specific presentation in the course by navigating to the timestamp indicated below.

0:00:00
Introduction

0:03:10
Beyond the hypnogram: Alternative representations of sleep structure
Merel van Gilst (Netherlands)

0:20:30
Unlocking sleep's secrets: AI-driven insights into brain health
Wolfgang Ganglberger (United States)

0:36:20
Sleep Monitoring Beyond the Lab: Opportunities and Limits of AI and Consumer Wearables Compared to PSG
Manuel Schabus (Austria)

0:57:45
AI-based tools for improving diagnosis and treatment of obstructive sleep apnea
Gabriel Natan Pires (Brazil)

1:16:00
From code to care: Implementing AI in sleep medicine devices and practices
Jon Agustsson (Iceland)

Summary

The field of sleep medicine is uniquely positioned to benefit from advancements in artificial intelligence (AI) algorithms and technologies. AI tools, particularly those designed for polysomnography analysis (such as automated sleep stage scoring) are already well-validated and are being incorporated into clinical practice. However, the potential of AI in sleep medicine extends far beyond automated scoring, promising transformative advancements in diagnosis, treatment, and patient care.
This symposium aims to provide a comprehensive overview of the broader applications of AI in sleep medicine, focusing on two primary areas. First, AI could offer new ways to interpret sleep recordings either by providing novel ways of interpreting sleep dynamics, or by uncovering patterns that may predict patients' future health. Second, AI has the capacity to assist clinicians in developing personalized treatment plans, a critical need in managing highly prevalent and heterogeneous sleep disorders such as insomnia and sleep apnea. While these applications are highly promising, the implementation of AI in these areas is not without challenges: regulatory aspects must be carefully navigated to ensure the safe and effective use of AI in clinical practice. These critical issues will also be addressed at the end of the symposium.

After an introduction of the chairs Dr. Cesari and Dr. Schabus, the symposium will feature five talks:

Dr. Van Gilst will introduce how AI algorithms allow to develop alternative representations of sleep structure, based on uncertainty measures. She will show their usefulness to identify and characterize several sleep disorders.
Dr. Ganglberger will talk about the potential of AI algorithms to analyze sleep recordings for predicting the ageing process and for predicting future adverse outcomes, including cardiovascular and neurodegenerative diseases.
Dr. Schabus will show how AI and mobile technologies can revolutionize digital behavioral therapy for insomnia patients, by providing personalized treatments based on both subjective and objective home recordings.
Dr. Pires will provide an overview of the applications of AI for personalized diagnosis and treatment of obstructive sleep apnea patients.
Dr. Agustsson will conclude the symposium by discussing the regulatory challenges for implementing these AI applications in the clinical routine. In particular, he will discuss the novel European AI act and its consequences in the field of AI in sleep medicine.
By bringing together clinicians, sleep researchers and engineers, we aim to highlight the transformative potential of AI in sleep medicine beyond automatic scoring, while fostering dialogue on the practical steps needed to bring these innovations to fruition.

Learning Objectives:

Upon completion of this CME activity, participants will be able to:
• Provide an overview of alternative representations of sleep structure and Introduce the hypnodensity representation and its derived metrics
• Show applications where hypnodensity and other representations provide more insight in sleep disorders not captured by the traditional hypnogram
• Learn how AI models process rich sleep data, including EEG dynamics
• Discover how sleep biomarkers predict healthy aging, cognitive health and disease risks
• Explore multimodal AI integration for advancing diagnostics and personalized care
• Interpret the potential of longitudinal sleep monitoring for early detection
• Learn how sleep can be classified using solely heart data
• Identify why subjective as well as accurate objective measures are needed in sleep research
• Define how to use objective sensor data in the therapy and/or optimization of sleep and Show the potentials of such approaches with respect to “precision medicine
• Recognize the use of AI for OSA diagnosis, using either regular polysomnographies or portable/wearable devices
• Learn how AI may lead to more tailored OSA treatment
• Describe AI-based approaches for the miniaturization and portabilization of diagnostic devices and implementation of indirect respiratory events detection
• Learn about AI algorithms as medical devices or clinical decision support systems, the process of releasing AI medical devices in the USA and EU, and how evolving regulations balance innovation with safety. • • •Explore the critical role of interdisciplinary collaboration in developing user-centric, validated AI solutions for clinical practice