PERISCOPE matches doctors in spotting potential post-surgery complications.
Leiden University Medical Center (LUMC) has developed a pioneering AI model, PERISCOPE, designed to predict infection risk in patients following surgery. PERISCOPE will become a standard tool at LUMC, with full implementation expected by mid-2026.
Based on data from over 250,000 surgical procedures, the tool provides a personalised risk assessment within seven to thirty days post-operation, helping healthcare providers intervene earlier and reduce complications.
The AI model, developed by PhD researcher Siri van der Meijden, uses pseudonymised patient data including medical history, vital signs and existing conditions to identify those most at risk.
During testing, PERISCOPE performed as well as experienced doctors and outperformed less experienced ones, making it a valuable decision-support tool. Once fully adopted, the tool is expected to save time, improve patient outcomes, and potentially predict other complications.
Rather than replace clinicians, it complements their judgement by offering a clear, visual dashboard of infection risk levels. Integration into hospital systems remains a challenge, but preparations are underway.
Van der Meijden continues to develop the model to expand its predictive capabilities and ensure long-term impact not only in the Netherlands, but globally.
Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!