In a groundbreaking move to combat colorectal cancer, the DECADE research project becomes the first to employ swarm learning (SL), a cutting-edge form of artificial intelligence (AI), in cancer research. This pioneering initiative is funded by the German Cancer Aid with approximately €1.5 million and collaboratively led by several German university hospitals.
Bowel cancer, one of the most prevalent and lethal forms of cancer in Germany, affects around 58,000 people every year. Although early detection can lead to effective treatments, many cases continue to pose substantial challenges in diagnosis and prognosis.
The DECADE project (Decentralised artificial intelligence for diagnosis, prognostication and response prediction in colorectal cancer) aims to leverage AI and SL to improve the detection, prognosis, and individualised treatment of colorectal cancer patients.
The use of AI in the medical field has gained traction for its capability to analyse large data sets and discern specific patterns, facilitating improved disease predictions and personalised diagnoses. However, the adoption of AI tools in everyday clinical practice has been slow due to legal and ethical constraints on data exchange between hospitals, particularly in Germany.
“Swarm learning provides a promising solution to these issues,” explained Professor Jakob N. Kather, project leader and professor of clinical artificial Intelligence at the Else Kröner Fresenius Centre for Digital Health at TU Dresden and the University Hospital Dresden.
Swarm learning is an innovative form of machine learning that enables multiple institutions to collaboratively train medical AI models without data sharing. It circumvents privacy laws and ethical concerns by using a blockchain for coordinating and integrating models, thus eliminating the necessity of a central authority.
The DECADE project seeks to pioneer the use of SL-based AI technology in addressing clinical challenges associated with colorectal cancer. “This method allows us to harness the advantages of collaboration and knowledge transfer between research institutions without breaching privacy regulations,” noted Professor Tom Lüdde, director of the Department of Gastroenterology, Hepatology, and Infectious Diseases at the University Hospital Düsseldorf.
DECADE partners intend to leverage SL to develop AI algorithms for diagnosing and subtyping colorectal cancer and predicting disease progression. This trailblazing use of SL in medicine may serve as a benchmark for all AI systems in the healthcare sector.
The research project DECADE is sponsored by the German Cancer Aid for a span of three years (2023-2026), partnering with university hospitals in Bonn, Dresden, Düsseldorf, Heidelberg, and Mainz. Its primary objective is to utilise AI and SL to enhance the treatment procedures for colorectal cancer patients.
The project involves various institutes including the Else Kröner Fresenius Center for Digital Health, TU Dresden, University Hospital Dresden, University Hospital Heidelberg, Institute of Pathology, Department of Applied Tumor Biology, University Hospital Mainz, Institute of Pathology, University Hospital Bonn, Medical Clinic and Polyclinic 1, National Center for Hereditary Tumor Syndromes, and University Hospital Düsseldorf, Department of Gastroenterology, Hepatology, and Infectious Diseases.
