Healthcare is rapidly transitioning from a fragmented system of specialized fields to a more holistic and multidisciplinary approach. However, combining information from various aspects of health, including physical, mental, and behavioral data, into a unified and actionable understanding remains a significant challenge. At Cyber3Lab, we fully exploit AI’s ability to detect patterns in large datasets. This allows to extract comprehensive health insights, enabling holistic personalized care.

This approach is demonstrated in our “Power Up Wellbeing! (PUW!)” project, where AI is used to analyse wearable data, psychological questionnaires, and emotional reports to monitor the wellbeing of secondary school students. This work has been presented at the Velon Velov conference in Rotterdam and published as a preprint at https://osf.io/preprints/psyarxiv/436cz_v1.

Beyond pattern recognition, AI can predict life-saving interventions. It has been claimed that algorithms analyzing heart rate sensor data can forecast heart failure up to five days in advance (see https://www.linkedin.com/pulse/ai-driven-remote-monitoring-tech-behind-real-time-health-xveqc/), empowering healthcare professionals to deliver timely treatment. This technology offers immense potential for supporting at-risk patients, which we aim to further explore.

Cyber3Lab is also pioneering AI-powered recommendation systems to motivate active, healthy lifestyles. In collaboration with researchers from Ghent University, we personalize activity recommendations using wearable data, GPS, weather, and personal information (see https://link.springer.com/article/10.1007/s11257-025-09427-3). This personalized approach has proven effective in promoting long-term behavioral change, which is crucial for preventive healthcare. We are still open for partners from industry who want to learn and explore the potential of recommendation systems with us.

Perhaps the most critical aspect of AI in healthcare is data governance and privacy. While AI can do powerful things with large amounts of data, healthcare deals with highly sensitive and personal data. At Cyber3Lab, we are specializing in the use of Solid data pods: personal data vaults that give patients complete control over their information while ensuring seamless interoperability among healthcare professionals (https://solidproject.org/).

Another interesting avenue with respect to data privacy is the recently proposed LLM-Anonymizer (https://ai.nejm.org/doi/full/10.1056/AIdbp2400537) which helps healthcare organizations protect patient privacy by removing sensitive information from medical documents before sharing or analysis. By running locally on hospital systems, it safeguards patient data and supports compliance with privacy regulations like HIPAA and GDPR, enabling secure research and AI development.