When AI is used in clinical practice, data often needs to be combined across many digital systems, where the challenge becomes clinical implementation and harmonization to handle missing data. At Rigshospitalet, researchers, including Chief Physician and Researcher Carsten Utoft Niemann, have developed an algorithm, Chronic Lymphocytic Leukemia–Treatment Infection Model (CLL-TIM), designed to predict the risk of infection and the need for treatment in patients with chronic lymphocytic leukemia (CLL).
The algorithm is fully data-driven and is now implemented in the electronic patient record system, Sundhedsplatformen, as well as in an international clinical trial. This implementation serves as a “cookbook” for scaling and implementing other data-driven decision support tools in electronic record systems.
The study shows a way to develop transparent data-driven medical practice, where the researcher and clinician remain in the driver’s seat, improve patient care, and save time on manual data handling.
This is a step towards the medical technology of the future and its practical applications. Carsten Utoft Niemann envisions a future where thousands of models like CLL-TIM can improve patient treatment. He calls for changes in practice regarding processes in Sundhedsplatformen and handling data from it.
Carsten Utoft Niemann - Hematologist, Chief Physician and Associate Professor - Rigshospitalet and University of Copenhagen
Oplægget er optaget på Digital Tech Summit 2024, der fandt sted den 30. og 31. oktober. Digital Tech Summit er en årlig messe og konference, der sætter universiteter og virksomheder sammen om AI og digitalisering. Eventen udbydes af Ingeniøren og tiltrækker hvert år omkring 5.000 deltagere.
