Cardiac hemodynamics can be computed from medical imaging data, and results could potentially aid in cardiac diagnosis and treatment optimization. However, simulations are often based on simplified geometries, ignoring features such as papillary muscles and trabeculae due to their complex shape, limitations in image acquisitions, and challenges in computational modeling. This severely hampers the use of computational fluid dynamics in clinical practice. The overall aim of this study was to develop a novel numerical framework that incorporated these geometrical features. The model included the left atrium, ventricle, ascending aorta, and heart valves. The framework used image registration to obtain patient-specific wall motion, automatic remeshing to handle topological changes due to the complex trabeculae motion, and a fast interpolation routine to obtain intermediate meshes during the simulations. Velocity fields and residence time were evaluated, and they indicated that papillary muscles and trabeculae strongly interacted with the blood, which could not be observed in a simplified model. The framework resulted in a model with outstanding geometrical detail, demonstrating the feasibility as well as the importance of a framework that is capable of simulating blood flow in physiologically realistic hearts.

Purpose To investigate four-dimensional (4D) flow CT for the assessment of intracardiac blood flow patterns as compared with 4D flow MRI. Materials and Methods This prospective study acquired coronary CT angiography and 4D flow MRI data between February and December 2016 in a cohort of 12 participants (age range, 36-74 years; mean age, 57 years; seven men [age range, 36-74 years; mean age, 57 years] and five women [age range, 52-73 years; mean age, 64 years]). Flow simulations based solely on CT-derived cardiac anatomy were assessed together with 4D flow MRI measurements. Flow patterns, flow rates, stroke volume, kinetic energy, and flow components were quantified for both techniques and were compared by using linear regression. Results Cardiac flow patterns obtained by using 4D flow CT were qualitatively similar to 4D flow MRI measurements, as graded by three independent observers. The Cohen κ score was used to assess intraobserver variability (0.83, 0.79, and 0.70) and a paired Wilcoxon rank-sum test showed no significant change (P > .05) between gradings. Peak flow rate and stroke volumes between 4D flow MRI measurements and 4D flow CT measurements had high correlation (r = 0.98 and r = 0.81, respectively; P < .05 for both). Integrated kinetic energy quantified at peak systole correlated well (r = 0.95, P < .05), while kinetic energy levels at early and late filling showed no correlation. Flow component analysis showed high correlation for the direct and residual components, respectively (r = 0.93, P < .05 and r = 0.87, P < .05), while the retained and delayed components showed no correlation. Conclusion Four-dimensional flow CT produced qualitatively and quantitatively similar intracardiac blood flow patterns compared with the current reference standard, four-dimensional flow MRI. © RSNA, 2018 Online supplemental material is available for this article.

Blood flow simulations are making their way into the clinic, and much attention is given to estimation of fractional flow reserve in coronary arteries. Intracardiac blood flow simulations also show promising results, and here the flow field is expected to depend on the pulmonary venous (PV) flow rates. In the absence of in vivo measurements, the distribution of the flow from the individual PVs is often unknown and typically assumed. Here, we performed intracardiac blood flow simulations based on time-resolved computed tomography on three patients, and investigated the effect of the distribution of PV flow rate on the flow field in the left atrium and ventricle. A design-of-experiment approach was used, where PV flow rates were varied in a systematic manner. In total 20 different simulations were performed per patient, and compared to in vivo 4D flow MRI measurements. Results were quantified by kinetic energy, mitral valve velocity profiles and root-mean-square errors of velocity. While large differences in atrial flow were found for varying PV inflow distributions, the effect on ventricular flow was negligible, due to a regularizing effect by mitral valve. Equal flow rate through all PVs most closely resembled in vivo measurements and is recommended in the absence of a priori knowledge.Electronic supplementary materialThe online version of this article (10.1007/s10439-018-02153-5) contains supplementary material, which is available to authorized users.

Objectives To evaluate an artificial intelligence (AI)-based, automatic coronary artery calcium (CAC) scoring software, using a semi-automatic software as a reference. Methods This observational study included 315 consecutive, non-contrast-enhanced calcium scoring computed tomography (CSCT) scans. A semi-automatic and an automatic software obtained the Agatston score (AS), the volume score (VS), the mass score (MS), and the number of calcified coronary lesions. Semi-automatic and automatic analysis time were registered, including a manual double-check of the automatic results. Statistical analyses were Spearman's rank correlation coefficient (⍴), intra-class correlation (ICC), Bland Altman plots, weighted kappa analysis (κ), and Wilcoxon signed-rank test. Results The correlation and agreement for the AS, VS, and MS were ⍴ = 0.935, 0.932, 0.934 (p < 0.001), and ICC = 0.996, 0.996, 0.991, respectively (p < 0.001). The correlation and agreement for the number of calcified lesions were ⍴ = 0.903 and ICC = 0.977 (p < 0.001), respectively. The Bland Altman mean difference and 1.96 SD upper and lower limits of agreements for the AS, VS, and MS were − 8.2 (− 115.1 to 98.2), − 7.4 (− 93.9 to 79.1), and − 3.8 (− 33.6 to 25.9), respectively. Agreement in risk category assignment was 89.5% and κ = 0.919 (p < 0.001). The median time for the semi-automatic and automatic method was 59 s (IQR 35-100) and 36 s (IQR 29-49), respectively (p < 0.001). Conclusions There was an excellent correlation and agreement between the automatic software and the semi-automatic software for three CAC scores and the number of calcified lesions. Risk category classification was accurate but showing an overestimation bias tendency. Also, the automatic method was less time-demanding. Key Points • Coronary artery calcium (CAC) scoring is an excellent candidate for artificial intelligence (AI) development in a clinical setting. • An AI-based, automatic software obtained CAC scores with excellent correlation and agreement compared with a conventional method but was less time-consuming.