Artificial Intelligence Based Therapy Support in Cardiology

Therapy planning and risk assessment for coarctation of the aorta and aortic valve disease.
Real-time prediction of patient-specific aortic hemodynamic parameters.
Based on standard clinical MRI data.

Cardiovascular diseases are the leading cause of death in Germany. The total cost for treatment of valve defects alone amounts to € 4.6 billion with 31,000 surgeries per year. Demand is expected to further increase in the future because of to the aging population. The physician in charge makes a therapeutic decision based on clinical guidelines and individual experience. However, guidelines are generic, not evidence-based, and do not adequately account for differences between patients.

Aided by innovative software, relevant patient-specific information derived from the patient’s data will be provided to the physicians to quickly, efficiently and safely determine the need, timing and type of individual therapy. The primary focus of this project is to develop digital therapy support for coarctation of the aorta, one of the most common congenital heart defect, as well as heart valve defects, the most common acquired heart disease.

The primary patient’s data provided by the digital therapy support is the individual anatomy of the patient, which is obtained by means of magnetic resonance imaging (MRI). Mathematical methods are used to characterize anatomies and to calculate flow conditions in the blood vessels with the aim of predicting the outcome of various surgical therapies. Since such computations cannot be performed efficiently online or on commercially available hardware, a novel method based on artificial intelligence is researched in this project. This should enable a practical and cost-effective integration into the clinical workflow, such that every patient can ultimately benefit from improved and more personalized treatment. The personalized therapy planning aims to shorten operation times, to avoid unnecessary therapies or follow-up surgeries, and thus to reduce the healthcare costs by up to 30%. The results of the project will also be used for other circulatory diseases.