Proefschrift

Healthcare Information System Engineering: AI Technologies and Open Source Approaches

Healthcare Information Systems (HIS) are essential for modern healthcare delivery, yet their development faces significant challenges including heterogeneous data formats, regulatory compliance, and the growing demand for AI-driven decision support.

Auteur: Z. Shen
Datum: 03 december 2025
Links: Thesis in Leiden Repository

This dissertation investigates how artificial intelligence technologies and open source principles can accelerate HIS engineering to address real-world clinical problems.The research addresses the central question: How can we employ artificial intelligence technologies -- such as machine learning algorithms, knowledge systems and natural language processing techniques -- based on open source principles to accelerate healthcare information system engineering in solving real-world clinical problems?We adopt a two-part methodology: first, developing practical HIS solutions to demonstrate AI applications in clinical settings; second, investigating how open source software can enhance HIS development. The first part presents STRIPA, a GDPR-compliant clinical decision support system for polypharmacy reviews; a lightweight API-based architecture for clinical NLP that reduces manual data entry; an automated NLP pipeline for extracting adverse drug reactions from regulatory documents; and a cloud-based framework for large-scale biomedical literature mining. The second part develops a reproducible methodology for systematically studying open source clinical software and introduces LOCATE, a web platform connecting open source clinical software with scientific literature.This work demonstrates that combining AI technologies with open source principles creates practical, scalable, and cost-effective solutions for HISdevelopment. All research artifacts (except the first clinical system) are publicly available on GitHub, contributing to the open source healthcare software ecosystem.