LUMC is investigating the use of AI in diagnosis of uterine cancer

Project leader and pathologist Tjalling Bosse keeps up with his research group has been working for years on improving the diagnosis of which patients should and should not receive additional treatments. These treatments are necessary for patients with a high risk of tumor recurrence.

Fortunately, a large proportion of patients with uterine cancer can be cured by surgically removing the uterus. But a major problem remains that it is difficult to predict which patients will benefit from additional treatments such as radiation or chemotherapy, ”says Bosse.

Improve diagnostics

The aim of the AIR-MEC study that is now being started is to improve diagnostics based on the assessment of the tumor with the help of artificial intelligence. Research by the Pathology and Radiotherapy departments has already shown that assessing molecular tumor properties improves the estimate. However, these new insights do not yet provide sufficient improvement in diagnostics.

“Although these recent insights into the genetic background of the tumors improve the risk assessment of tumor recurrence, we are not yet satisfied. The necessary tumor tests are not available all over the world and also involve additional costs. This means that this knowledge cannot be brought to the patient everywhere, ”says Bosse.

Artificial intelligence

The use of artificial intelligence, or AI, is a technique that is already used in several forms of cancer diagnosis deployed. Bosse’s team will now investigate whether this technology can also improve risk assessment for post-operative treatments for uterine cancer. “Recent examples show that computers are extremely capable of recognizing patterns in microscopy images and making predictions based on them,” says Bosse.

For the AIR-MEC, the Bosse team has the largest tissue bank in the world to use. This consists of tumor tissue from patients who participated in the PORTEC studies. These clinical studies have shaped the current treatment of uterine cancer worldwide. The unique collection has been built up by us and mapped out completely molecularly. The tumor images form the basis of the AIR-MEC study ”, Bosse explains.

The research team has received a grant of 400,000 euros from the Hanarth Fund for the AIR-MEC study. A fund dedicated to improving and advancing the use of artificial intelligence and machine learning to improve the diagnosis, treatment and outcome of cancer patients. The research is being carried out in collaboration with pathologist Viktor Koelzer from the University of Zurich, Jouke Dijkstra from the Radiology Department and Carien Creutzberg from the Radiotherapy Department.

Share on facebook
Share on pinterest
Share on twitter
Share on linkedin
Share on email


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.