Startseite › Forschung › Projekte › Reasoning with Conditionals

Decicsion Support in Medical Cancer Therapy

For the medicamentous therapy of cancer, a large variety of different active pharmaceutical ingredients is available. Typically, several of these ingredients are combined in a therapy and are applied to the patient according to a given temporal scheme. For determining a patients's therapy, several important aspects originating from different information resources have to be taken into account. In particular, this includes the current medical guidelines regarding the present cancer type, the line of treatment, the individual situation of the patient, and the molecular factors of the tumor.

The parameters leading to a defined therapy are rapidly changing for multiple reasons. Many new drugs representing new therapeutic strategies have come to the clinic during the last five years, and their number is constantly increasing. In addition, new molecular technologies facilitate the knowledge of individual properties of each individual tumor. These developments lead to a rapid diversification and individualization of tumor therapy. Many attempts are undertaken to manage this ongoing informational diversification. Over the last 15 years, the pharmaceutical department together with the clinic of Hematology and Medical Oncology of the St.-Johannes-Hospital Dortmund, Germany, has developed an electronic support system containing treatment plans for more than 2,300 individual treatment situations. These plans are provided together with all necessary information on co-medication, behavioral rules and explanations for the patient. In addition, the system documents all therapies of more than 40,000 therapeutical cycles with the information of actual application, dose modifications, and tumor-board decisions. In an ongoing project, this data together with advanced knowledge representation and reasoning methods from computer science is the basis of an approach for the development of a comprehensive Artificial Intelligence based tool to support decision making taking into account all available clinical as well as molecular information of each patient and his tumor.

This is a joint project with the St.-Johannes-Hospital Dortmund, Germany, and the TU Dortmund (Prof. Dr. Gabriele Kern-Isberner).