Revolutionary AI study for faster colon cancer diagnosis started!

Revolutionary AI study for faster colon cancer diagnosis started!

An international, interdisciplinary research team led by Prof. Jakob N. Kather has developed an innovative AI model for analyzing colon cancer. This new technology could play a crucial role in the future of diagnostics. As the Technical University of Dresden reports, the model was developed at the Else Kröner Fresenius Center (EKFZ) for Digital Health at TUD.

The system detects both genetic and tissue changes by analyzing tissue section images. The aim of this development is a faster and more cost-effective diagnosis of colon cancer. The study analyzed nearly 2,000 digitized tissue sections from seven independent cohorts in Europe and the United States. These samples included not only tissue but also clinical, demographic and lifestyle data.

Innovative technologies in cancer research

The new “multi-target transformer model” stands out from previous studies, which mostly focused on genetic modification. It is capable of identifying many biomarkers simultaneously, including clinically ignored mutations. This particularly includes microsatellite instability (MSI), which is considered an important biomarker for immunotherapies. In testing, the model showed impressive prediction results for biomarkers such as BRAF and RNF43 mutations.

Experienced physicians brought pathological expertise into the assessment of tissue changes. The results of this important study were published in the renowned journal “The Lancet Digital Health”. This interdisciplinary collaboration spanned several leading research institutions in Europe and the USA, including the University of Augsburg, the NCT Heidelberg, the Fred Hutchinson Cancer Center as well as the Medical University of Vienna and the Mayo Clinic.

Outlook for future developments

The research team plans to transfer the newly developed approach to other types of cancer. In addition, other relevant research projects have also been launched to understand cancer metastasis. The DECIPHER-M project, which is also led by Prof. Kather, aims to use AI to decipher the mechanisms of cancer metastasis. This could enable crucial advances in the treatment of cancer metastases, which are considered one of the biggest challenges in cancer medicine German Cancer Research Center (DKFZ) notes.

Starting in March 2025, the five-year project will promote interdisciplinary collaboration between partners from Aachen, Dresden, Essen, Heidelberg, Mainz and Munich. The goal is a more precise prediction of the individual risk of metastasis and the development of personalized therapeutic approaches. Combinations of tissue samples, X-rays and MRI images as well as genetic information are used.

In addition to technological development, there will also be more intensive discussions about colorectal cancer prevention and treatment. Several symposiums on this topic took place in March 2023, organized by LMU Hospital and the Network Against Colon Cancer e.V. These events highlighted the current status of national colon cancer prevention as well as future-oriented prevention approaches.

These advances in cancer research could ultimately not only sustainably improve diagnostics, but also the quality of treatment and contribute to significantly reducing the mortality rate from colorectal cancer.