NEWS Umeå University is one of 79 partners from 14 EU member states in the four-year EUCAIM project. The goal is an infrastructure for medical cancer images with over 60 million anonymized X-ray examinations and clinical data from cancer patients, accessible to clinicians, researchers and innovators across the EU.
At its core is the investment Europe's Beating Cancer Plan, which encompasses a total of 4 billion euros. EUCAIM, the European Cancer Imaging Initiative, is flagship project 2 of several in this work, with the goal of collecting and making available at least fifty different AI tools, intelligent algorithms and clinical prediction models in the field of cancer within four years. With this, it is possible to achieve more accurate and faster clinical decision-making, diagnostics and treatment as well as precision medicine based on each patient's conditions.
The Swedish participation is coordinated by Katrine Riklund, professor and senior physician at the Department of Radiation Sciences, Umeå University. In Sweden, researchers from Region Västerbotten, Uppsala (SciLifeLab) and Linköping University as well as Karolinska Institutet also participate. From Umeå participates, among others the PREDICT infrastructure, a collaboration between Umeå University and Region Västerbotten with the goal of developing the use of biobank samples.
– There is a lot of data available in various data warehouses, but currently they are fragmented and therefore difficult to use in the development of, for example, algorithms for AI in precision medicine. The idea with EUCAIM is to create both a central and a distributed infrastructure, to be able to use AI for different issues. Images and other data may remain in different hospitals, but everyone has access to the description of all data, and this in a GDPR-safe way. To create this, we need, among other things, to build secure infrastructures for data sharing and solve several legal challenges, says Katrine Riklund.
She is responsible for Work Package 7, a large assignment that includes just over 1,300 person-months.
– We will build and fill a platform with data, i.e. X-ray images, pathology, information on protein expression and gene arrays, so-called "omics", as well as clinical parameters. The main goal of Work Package 7 is to, together with the users, create conditions so that the platform and the developed tools and algorithms can be used for AI development and validation in clinical environments. The aim is that the use of AI in clinical operations increases to the benefit of the patient.
In connection with the start of the project, it was noted that Sweden is involved in four projects (European Genomic Data Infrastructure, EDIH Health Data Sweden, EUCAIM and Testing and Experimentation Facility for Health) in health that are financed by the EU's DIGITAL call. A joint preliminary study has contributed solutions that facilitate the use of health data. In this collaboration, the focus is on ethics and legal aspects, technical infrastructure and automatic communication between computer systems.
– If we can solve some of these questions, then we have come a long way. WE have a lot to learn from each other and also reduce duplication of effort through collaboration. A lot of the project is about coordinating resources. For Umeå's part, I believe that the project will have great significance for the use of AI in the future.
The scope is significant in the project and everything from computer science to clinical medicine and companies are involved.
– We will also invite more parties in an upcoming tender. WE also need more employees locally in Umeå. Anyone who is interested is welcome to contact me, says Katrine Riklund.
For Sweden, the project generates SEK 12.4 million in research funds from the EU, with SEK 2.6 million going to Region Västerbotten and SEK 4.2 million to Umeå University.
EUCAIM is financed with EUR 36 million.
– In Sweden, Vinnova contributes approximately SEK 13 million for the Swedish co-financing in EUCAIM, which we are very happy about, says Katrine Riklund.
