Using AI to Manage Stress and Improve Working Conditions

Computational intelligence (CI) is transforming eHealth by enabling intelligent, adaptive and data-driven solutions to complex healthcare challenges. Machine learning, deep learning, evolutionary computation and fuzzy systems can support personalized care, early disease detection, remote patient monitoring and clinical decision-making. By integrating heterogeneous data from electronic health records, wearable sensors, medical images and mobile health platforms, CI-based applications generate actionable insights that can improve prevention, diagnosis, treatment and long-term disease management. These advances can enhance patients’ access to healthcare, strengthen their autonomy and enable more timely and precise interventions, while improving the efficiency and quality of healthcare delivery.

The Computational Biomedicine Laboratory (CBML) of the University of Piraeus has developed integrated homecare solutions that combine WebRTC-based communication, wearable devices, assistive environments and intelligent data-processing modules. These systems support people with chronic conditions and older adults in living independently while enabling the continuous assessment of their health status. Their capabilities have also been extended beyond conventional physiological monitoring through the incorporation of affective computing. The recognition of emotional and behavioral changes can facilitate the early identification of potentially harmful situations, given the close relationship between emotional state, physical health, cognitive functioning, behavior and overall quality of life.

In parallel, CBML investigates explainable artificial intelligence in medical image analysis, focusing on the evaluation and improvement of established explanation methods. Visual explanations that identify the image regions influencing a model’s predictions are particularly important in clinical decision-support systems. Explainability-by-design and post-hoc interpretation methods can improve transparency, support clinical validation and strengthen healthcare professionals’ trust in AI-assisted decisions.

This talk will present the application of these technologies across a range of use cases, including holistic health management, chronic disease management, independent living, COVID-19 monitoring, mental healthcare, stress management and medical image analysis. It will also critically discuss their capabilities, limitations and implementation challenges, particularly those related to privacy, interoperability, ethics, explainability and clinical reliability. Overall, the convergence of computational intelligence and eHealth offers a promising pathway towards proactive, trustworthy, personalized and sustainable healthcare, with the ultimate goal of improving quality of life.