Each year, about 22,000 individuals in the EU are diagnosed with a glioma, a malignant type of brain tumor. Median survival time after diagnosis can be as short as 15 months. A new consortium combining expertise within brain tumors, machine learning, and data security will be looking for new diagnostic and treatment methods.
The consortium consists of Umeå University and Umeå University Hospital (Region Västerbotten, Sweden), Heights.ai (The Netherlands), and PSE Data Security (Switzerland). The consortium members signed their collaboration agreement on December 28 at a meeting in Stockholm. The collaboration will leverage anomaly detection models (used for finding "a needle in the haystack") to improve the early detection of gliomas.
We have already seen that these models can be successful in other industries, such as with the detection of money laundering in the financial industry."
Wilfred de Graaf, partner at Heights.ai
Umeå University and the health care provider Region Västerbotten have amassed a unique set of health data and samples from 140,000 individuals over the past 30 years. This serves as an important backbone for the new collaboration since AI models need datasets of sufficient size and quality.
"At Umeå University, we have a history of translating innovations into clinical applications. The Nobel prizewinning CRISPR/CAS method was discovered here and is now in clinical trials at our University Hospital," says Beatrice Melin, Professor of Oncology at Umeå University.
"I am happy about developing this new collaboration with the private sector. I profoundly believe that stimulating innovations between different fields will speed discovery of novel methods for brain tumor detection."
Quantum biology's new frontier: Tryptophan networks and brain disease defense