FLORENCE: Federated Learning on Cancer Data

Background

Colon and rectal cancer is the second most common cancer in Norway when looking at both sexes. In 2021, 4,550 men and women were diagnosed with colon and rectal cancer. 1 in 4 bowel cancer patients experience complications after surgery, and the same proportion experience a recurrence of the cancer within 3 years. 

Better utilization of register and hospital data can give clinicians and patients a better basis for treatment and for the prevention of complications after surgery. Mapping/assigning data to the OMOP data model makes it possible to analyze data across countries without the data itself being moved across borders.

Purpose

In this project, we want to lay the foundations for an IT tool (decision support tool) that can be used by specialists to help with treatment choices for bowel cancer patients.  

The project is a collaboration with the Center for Surgical Science (CSS) at Zealand University Hospital in Denmark, and by using hospital and register data from both countries, we will form the basis for a new decision support tool. In addition, there is collaboration with Lund University in Sweden, Computerome at DTU and the unit for research projects at Zealand University Hospital. 

In order to create a decision support tool that uses basic data from several countries, it is important that the data on which the analyzes are based are comparable. This means that the data must be standardized and harmonized across the countries before they can be compared. To achieve this, we will use a generic data model called OMOP-CDM from Observational Health Data Sciences and Informatics (OHDSI). The project will also use synthetic data for testing and developing algorithms that can be regularly improved via federated learning. 

We have received funding for the project together with the project partners from Interreg Øresund-Kattegat-Skagerak (ØKS). Interreg ØKS is a European regional development fund that supports projects that focus on innovation, a green transition, transport or a borderless labor market.

Privacy

Privacy is i.a. safeguarded by using a common data model so that algorithms can be shared instead of health information. The local data at a hospital/register will be used to train each version of the model, or the decision support tool in this case. Changes to the model between each site can be shared, and in that way all treatment sites will have an equally good decision-making tool.

See informational video about cooperation in the FLORENCE project.