Data Science Projects

I dag oplever minimum én ud af fire patienter komplikationer efter tarmkræft-kirurgi, som medfører genindlæggelse, varige mén og i nogle tilfælde død. Et nyt forskningsprojekt, kaldet FLORENCE, skal forbedre diagnosticering, prognose og behandling af patienter med tarmkræft ved at anvende kunstig intelligens (Artificial Intelligence).

In addition to AI, the FLORENCE project uses the OMOP Common Data Model,  which is one of the leading approaches to creating data infrastructures that facilitate the use of personalised medicine in medical care.

FLORENCE projektet anvender OMOP Common Data modellen, som er en førende tilgang til at skabe datainfrastrukturer, der fremmer anvendelsen af personlig medicin i sundhedsvæsenet (dvs. skræddersyet behandling af den enkelte patient). Via det, som kaldes federeret læring, vil projektet som noget helt nyt på globalt plan, koble AI-modellen direkte til klinikken på hospitalet. Projektet vil på længere sigt være med til at skabe en international best practice for implementering af registerdatabaserede AI-modeller i klinisk praksis.

The project’s lead partner is the Center for Surgical Science and the Research Department at the University Hospital of Zealand in Køge. Partners are Oslo University Hospital (respectively the Oncological Pelvic Surgery Unit and the Cancer Registry of Norway), Lund University og DTU. The Netherlands Cancer Registry (IKNL) is also participating in the project.

FLORENCE er en forkortelse for centrale aktiviteter i projektet (dvs. federeret læring med OMOP-modellering af sundhedsdata for at forbedre tarmkræftbehandling i Norden).

One of the main targets of Personalized Medicine is to refine the stratification of a single patient aiming to provide improved diagnosis, prognosis, and treatment for the specific individual. This can be achieved by including not only a detailed account of the patient’s health record, but also similar data from population-wide databases and clinical projects. These multiple data sources can be collected in a single and common data model to serve as a powerful platform. The model will include the necessary data science domains such as machine learning and aid the clinician in the decision-making process. To facilitate the implementation, real-time data from the electronic health report of the patient is needed. We will, in collaboration with data scientists and medical and biological professionals, establish the necessary infrastructure to promote this model in a clinical use case and finally, expand it within other medical fields. 

The project is funded by the Novo Nordisk Foundation.