The DECADE research project was the first to use swarm learning in cancer research. German Cancer Aid is funding some 1.5 million euros for joint projects of several university hospitals.
Bowel cancer is one of the most common and deadly cancers in Germany. About 58,000 people are diagnosed with the disease each year. If detected early, colorectal cancer is easily curable. However, despite significant advances in screening and treatment, physicians still face diagnostic and prognostic challenges. This is where DECADE (Decentralized Artificial Intelligence for Colorectal Cancer Diagnosis, Prognosis, and Response Prediction) comes in. Several German university hospitals are collaborating to see how artificial intelligence (AI) and swarm learning (SL) can be used to significantly improve diagnosis, prognosis and response prediction in colorectal cancer. Researching. Care and treatment of patients with both early and advanced stage colorectal cancer. AI is already able to analyze large amounts of data and recognize specific patterns. The insights gained from this will help us better predict the course of the disease and make more individualized diagnoses. The aim of this research project is to use AI and SL to significantly improve the treatment of colorectal cancer patients.
Professor Jakob N. Kather, project leader and professor of clinical artificial intelligence at the Else Kroner Fresenius Digital Health Center at Dresden University of Technology and Dresden University Hospital, said at the start of the project: It is used with hesitation in routine clinical practice. One reason for this is that, especially in Germany, data exchange between hospitals is severely restricted by legal and ethical hurdles. One solution is swarm learning. Swarm learning allows multiple institutions to jointly train medical AI models by exchanging data. By using distributed artificial intelligence and swarm learning, we hope to improve diagnosis, prognosis, and treatment planning for colorectal cancer patients. ”
Train AI with decentralized patient data
In cancer research, privacy laws and ethical hurdles lead to the sharing of confidential patient data between different research institutions, even though many patients in principle agree that their data be used for research purposes. is becoming difficult. Swarm learning makes it easier to meet privacy requirements. Swarm learning is a special form of machine learning that trains models without exchanging actual data between participants. Coordination and coupling of models is done via blockchain, so no central instance is required. Building on this methodology, the DECADE project uses SL-based AI technology to solve real-world clinical problems related to colorectal cancer.
Legal requirements to protect sensitive health data are high. This innovative swarm learning method allows us to realize the benefits of collaboration and knowledge transfer between different research institutions without violating privacy regulations. In this way, AI models in cancer research can be further developed and improved to enable better diagnostics, prognosis and personalized treatment approaches for cancer patients. ”
Prof. Tom Rudde, Head of Gastroenterology, Hepatology and Infectious Diseases, Düsseldorf University Hospital
Project partners will use SL to develop AI algorithms for diagnosing and subtyping colorectal cancer and predicting disease progression. In doing so, they are setting a precedent for the medical use of SL to serve as a template for any AI system in the medical field. Ultimately, more powerful AI systems could help doctors detect colorectal cancer earlier and treat it more effectively. This has the potential to support medical staff and improve care and treatment for colorectal cancer patients.
Background information:
The research project DECADE (Decentralized Artificial Intelligence for Diagnosis, Prognosis and Response Prediction in Colorectal Cancer) is funded by the German Cancer Aid for approximately €1.5 million over three years (2023-2026) . Project partners are the university hospitals of Bonn, Dresden, Düsseldorf, Heidelberg and Mainz. The aim of the research project is to use artificial intelligence and swarm learning to improve treatment of colorectal cancer patients.
Project partner:
- Else Kröner Fresenius Center for Digital Health, Dresden University of Technology, Dresden University Hospital
- Heidelberg University Hospital, Institute of Pathology, Department of Applied Tumor Biology
- Institute of Pathology, Mainz University Hospital
- Bonn University Hospital, Clinic and Polyclinic 1, National Center for Hereditary Oncology Syndrome
- Düsseldorf University Hospital, Department of Gastroenterology, Department of Hepatology, Department of Infectious Diseases
sauce:
Dresden University of Technology
