Training Data for Predicting Hemodynamics in Aortic Stenosis Patients

This dataset has been derived from computed-tomography based CFD simulations of the human aorta in patients with aortic valve stenosis. A total of 309 cases is presented, which are based on 103 patients. Each patient was simulated with three different flow rates (e.g. case23_01, cas23_02, case_03). _01 is the the baseline flow, _02 and _03 are ±25% of the baseline flow.The data files contain static pressure and wall-shear-stress along the aortic centerline (output sequence) as well as centerline coordinates, cross section area/shape and flow rate (input sequence). The data can be used to train an artificial neural network to predict pressure and wall-shear-stress from geometric data of patients with aortic stenosis.Two files are provided:One contains the input and output sequencesThe other contains the cross-sections in MATLAB cell format. Each cell holds an image stack with each image representing one cross-section for that case. Each cell in the file corresponds to the case in the sequence files. E.g. cell no. 22 in the Matlab file contains the cross sections for case case22_01 case22_02 and 22_03.