About Digitalisation and AI, Data Protection, Data Exchange, Data Mining—Legal Costraints/Challenges Concerning Sleep Medicine
Dr. Bernd Feige
Physicist and electrophysiologist who graduated from University of Münster, Germany, with a doctorate in magnetoencephalography. For the last thirty years he has been at Freiburg University, Germany, working on EEG, MRI, sleep and insomnia.
Digitalisation and AI in sleep medicine
This paper explores the legal and practical challenges arising from the growing use of digital tools and artificial intelligence (AI) in sleep medicine, particularly concerning data protection, data sharing, and algorithm transparency. It highlights the tension between the clinical potential of these technologies and the responsibility to safeguard patient privacy and trust.
With the digitalisation of sleep diagnostics, especially through polysomnography, the field now generates substantial volumes of data. AI-based systems are increasingly used to support sleep staging, assist with diagnosis, and advance research. However, many of these systems are developed using proprietary methods, raising concerns about data access, ownership, and reproducibility.
A key focus of the paper is the protection of sensitive health data under the General Data Protection Regulation (GDPR). Effective use of AI depends on structured and annotated datasets, often drawn from patient records. Ensuring compliance requires clear consent frameworks, robust data security, and well-defined governance protocols, particularly when data is shared across institutions or with external technology providers.
The discussion also addresses potential biases in algorithmic models trained on unbalanced or unrepresentative datasets. Variations in data collection methods and the opaque nature of many neural networks pose additional challenges to transparency and trust. Despite these limitations, the article outlines how standardised data formats, federated learning approaches, and explainable AI could support more ethical and robust applications of these tools in sleep research and care.
In conclusion, the paper calls for a harmonised legal and ethical framework that enables innovation while preserving clinician oversight and patient rights. AI should remain a tool to support, not replace, clinical decision-making, ensuring its integration strengthens, rather than undermines, the practice of sleep medicine.
Watch the interview with Dr. Bernd Feige, where he discusses the motivation behind the paper and its key messages:
Links to Paper:
Feige, B., Benz, F., Dressle, R.J. and Riemann, D. (2025), About Digitalisation and AI, Data Protection, Data Exchange, Data Mining—Legal Constraints/Challenges Concerning Sleep Medicine. J Sleep Res
Recent publications from ESRS members
- Palagini, L., et al (2025). Clinical practice guidelines for switching or deprescribing hypnotic medications for chronic insomnia: Results of European neuropsychopharmacology and sleep expert’s consensus group. Sleep Medicine
- Ibrahim A, Högl B, Stefani A. (2025) Sleep as the Foundation of Brain Health. Semin Neurol.
- Gramkow, M.H., Brink-Kjær, A., Clemmensen, F.K. et al. (2025). Diagnostic performance of actigraphy in Alzheimer’s disease using a machine learning classifier – a cross-sectional memory clinic study. Alz Res Therapy
- Espie CA, Grandner MA, Drake CL. (2025). PRO: CBT-I for insomnia treatment can be delivered by a range of healthcare providers. Sleep
- Picard-Deland C, Cesari M, Stefani A, Maranci JB, Hogl B, Arnulf I. (2025) The Future of Parasomnias. J Sleep Res.
- Markovic A, Veen D, Hamann C, Adorjan K, Kaess M, Tuura O’Gorman R, Tarokh L. (2025) Joint Heritability of Sleep EEG Spindle Activity and Thalamic Volume in Early Adolescence. J Neurosci.