Many factors must be taken into account when making medical decisions. In addition to medical expertise, previous test results and the patient's medical history also play an important role. Doctors use this information as a basis for making a diagnosis. In project C05, computer scientists at Bielefeld University are developing an intelligent, interactive system called ASCODI that supports medical experts in this complex task. The system helps to systematically compare different diagnostic hypotheses with the available medical evidence.

ASCODI not only answers questions, but also asks specific follow-up questions and makes recommendations. This collaboration between humans and the system enables experts to make informed diagnostic decisions and explain transparently why a particular diagnosis is the most plausible compared to other possibilities.

In the second funding phase of TRR 318, the system will be comprehensively tested and further developed. The focus will be on jointly building arguments for or against a diagnostic hypothesis and further refining its validity in an iterative, interactive process.

Research areas: Computer science