Entscheidungsunterstützung Bei Herzklappenerkrankungen Anhand Funktionaler Und Morphologischer Eigenschaften Mittels Methoden Der Mathematischen Modellierung Und Des Maschinellen Lernens

Using temporally resolved computed tomography images, the patient-specific geometry of the stenosed aortic valve during peak systole is reconstructed. Using this geometry, the projected area of the aortic valve's orifice (AVA) is calculated. Additionally, the left ventricular geometry is reconstructed for the complete heart cycle. Using the information of the left ventricular volume change, the patient-specific flow profile and peak-systolic flow (Q)is calculated. Using this information, the pressure gradient is calculated using a power law estimation of the form PG = a * AVA^b * Q^c. The model generation and parameter fit is described in [1].

To validate this model, retrospective data of patients receiving a catheter-based replacement of the aortic valve (TAVI) is collected. For those patients, CT images are already acquired for treatment planning and the invasive pressure measurements are performed during replacement of the aortic valve. Therefore, no additionally steps are required. Using the CT images the patient-specific aortic valve area ad flow rate are calculated. This information is then used for estimation of the pressure gradient using the power law model. The catheter-based pressure gradient is calculated as the difference between the peak-systolic pressure in the left ventricle and the ascending aorta before implantation of the prosthesis.

[1] Franke et al.; Towards improving the accuracy of aortic transvalvular pressure gradients: rethinking Bernoulli; Medical & Biological Engineering & Computing (2020); https://doi.org/10.1007/s11517-020-02186-w