PRISMA-AI-enabled Multimodal Decision-Support Integrating Clinical Notes and Biomarkers for Early Risk Stratification in Chronic Diseases

We have developed a TRL6 clinical-note NLP engine validated on large, real-world EHR datasets (ICU/ED; multilingual Greek/English), demonstrating consistent applicability across chronic-disease and multimorbidity contexts. Its prognostic risk outputs enable earlier and more effective management of chronic disease trajectories, with strong results in CKD and clear potential for neurological multimorbidity. Through PRECISEU, we aim to advance this NLP engine to TRL8 via deployment in multiple clinical regions and integration into routine workflows. Additionally, we have observed that adding biological biomarkers improves predictive performance. During the project, we will also integrate biomarker data from CKD and neurological patient cohorts to build a unified multimodal clinical decision-support companion. We seek hospitals with structured EHR notes (internal medicine, nephrology, neurology) and partners experienced in EHR integration to support deployment and validation of the platform in real clinical workflows.