PATH develops and evaluates AI-supported tools for use in clinical workflows covering areas such as diagnosis support, risk prediction, treatment planning, and patient monitoring.
These tools are trained using large and diverse datasets from multiple European hospitals to improve accuracy, reduce bias, and ensure reliability in different clinical environments.
Explainable AI techniques are used to make AI decisions transparent and understandable for clinicians, increasing trust and facilitating adoption in daily practice. These tools are intended to support healthcare professionals, not replace them. Medical decisions remain the responsibility of qualified clinicians.
The project also supports federated learning, allowing AI models to be trained on distributed data without moving sensitive patient information.