KITTU 2.0 - Prospective Multicenter Clinical Study to Evaluate AI-Supported Therapy Recommendations for Patients in Urological Oncology

Institution: Klinik und Poliklinik für Urologie und Kinderurologie der Universitätsmedizin Mainz
Applicant: Dr. Gregor Duwe

Funding line:

Translational Research

Multicenter evaluation of the successfully developed KITTU-prototype for optimizing evidence-based treatment recommendations in urological oncology for future translation into real-world patient care.

Project partner:

Deutsches Forschungszentrum für Künstliche Intelligenz (DFKI) - Forschungsbereich Smarte Daten & Wissensdienste
Klinik für Urologie des Universitätsklinikums Tübingen
Zentrum für Uro-Onkologie des Ordensklinikums Linz, Tumorzentrum Oberösterreich

Project description:

A tumor board (or tumor conference) is an interdisciplinary panel of experts in which physicians and specialists from various fields (e.g. oncology, surgery, pathology, radiology) discuss the optimal diagnosis and treatment for a patient's cancer. In the future, this process can be supported by AI tools. Following the successful development of the “KITTU” prototype, the AI decision-support tool of the project leaders is currently undergoing a prospective multicenter evaluation. KITTU aims to provide physicians with all relevant therapeutic options within the interdisciplinary urological-oncological tumor board framework to optimize evidence-based treatment recommendation. The prototype can generate an explainable recommendation with over 70% accuracy. Morever, depending on the targets accuracy rates of over 90% were already achieved. Further development and evaluation is mandatory and offers great potential for translating the AI decision-support tool into real-world patient care.

Further information here.

Project leader Dr. Gregor Duwe and Clinical Project Manager Dr. Verena Kauth, on behalf of the entire KITTU team

Print