Hyun Woo Goo

Coronary artery involvement in Kawasaki disease should be evaluated throughout all disease stages. Generally, multimodality imaging is necessary for both the acute and chronic phases of coronary artery abnormalities. This review summarizes current clinical indications and recent technical advances in coronary artery imaging in Kawasaki disease with an emphasis on the roles of computed tomography and magnetic resonance imaging. Echocardiography remains the first-line modality in all disease stages for evaluating coronary artery lesions and ventricular function. Recent technical developments in coronary computed tomography angiography can provide the accurate diagnosis of coronary artery lesions including stenosis and thrombus at substantially reduced radiation dose, supporting its selective use in serial follow-up. Cardiac magnetic resonance imaging offers the comprehensive evaluation of inducible myocardial ischemia, ventricular function, inflammation, and fibrosis. Invasive catheter angiography is reserved for complex or interventional cases. This review also addresses technical considerations and potential pitfalls for each imaging modality. Continued technical developments and their optimization with standardized imaging protocols and reports are expected to enhance diagnostic precision and eventually improve clinical outcomes in children with Kawasaki disease.

Artificial intelligence-based segmentation and quantification is a hot topic in cardiovascular imaging. However, the development of clinically useful artificial intelligence-based models for congenital heart disease is challenging owing to morphologic heterogeneity and limited data for training. Although cardiovascular computed tomography is increasingly used in evaluating patients with congenital heart disease, only a few reports have evaluated artificial intelligence-based segmentation algorithms for congenital heart disease. To be familiarized with this trend, we need to understand state-of-the-art segmentation techniques for cardiovascular digital twins as well as basic concept, current challenges, and potential solutions of artificial intelligence-based methods. Additionally, in this review, our learnings from the literature and our pilot study are summarized, and future directions are suggested.

Masaud Shah, Hyun Goo Woo. KOBRA Reports 2025;1:. https://doi.org/10.64066/kobrareports.25.0025

In addition to morphologic evaluation, cardiac computed tomography with a short scan time and a high spatial resolution may be performed to assess ventricular function in children with congenital heart disease, especially when cardiac magnetic resonance imaging is unavailable, inadequate, or contraindicated. Caution should be exercised over the computed tomography radiation dose because greater radiation doses are used to assess ventricular function than morphology. The three-dimensional threshold-based segmentation approach leads to higher accuracy and reproducibility of ventricular volumetry than the conventional approach, which frequently uses simplified contouring and two-dimensional disc summation. However, this approach is less frequently utilized in clinical practice than the conventional approach. Therefore, a step-by-step guide to cardiac computed tomography ventricular volumetry using three-dimensional threshold-based segmentation is presented for congenital heart disease. Graphical abstract

In conjunction with catheter angiography, high-resolution, isotropic, three-dimensional computed tomography mapping of major aortopulmonary collateral arteries is useful for delineating complex thoracic arterial anatomy and has been increasingly utilized in patients with pulmonary atresia and ventricular septal defect. The presence, confluence, and size of the central (mediastinal native) pulmonary artery; the presence and course of a patent ductus arteriosus; the number of collateral arteries; the number of lung segments supplied by each collateral artery; significant distal pulmonary artery stenoses; and isolated or dual blood supply in each lung segment can be evaluated using computed tomography mapping. The detailed mapping can substantially reduce procedural time for catheter angiography, which delivers a higher radiation dose than computed tomography does, and is useful for surgical planning. Furthermore, the effect of surgical treatments including unifocalization and rehabilitation can be monitored with computed tomography mapping. In addition to pulmonary atresia with ventricular septal defect, a major aortopulmonary collateral artery may be observed in other cardiac defects associated with pulmonary atresia, scimitar syndrome, and systemic arterial supply to normal lung. Lastly, the review briefly discusses promising future applications of computed tomography mapping focusing on quantification of arterial and lung volumes. Graphical Abstract

Background Optimization of localizer radiograph and bolus tracking doses is necessary, as their dose proportion may increase with a decreasing diagnostic scan dose in pediatric cardiothoracic computed tomography (CT). Objective To evaluate the radiation dose proportions of the localizer radiograph and bolus tracking in low-dose pediatric cardiothoracic CT. Materials and methods For low-dose pediatric cardiothoracic CT, a posteroanterior localizer radiograph was acquired with 80 kV, and 35 mA or 20 mA in 852 infants (agez-axis proportion, visibility of anatomic landmarks, and image noise. The over z-axis coverage proportion was correlated with the dose proportion of the localizer radiograph in both groups. Additionally, bolus tracking was performed in 1,015 children (≤2 years). The effects of the number of monitoring scan, dose-length product of the diagnostic scan, age, and water-equivalent area of the scanned patient body on the radiation dose proportion of bolus tracking were evaluated. Results The dose proportion of the localizer radiograph was significantly lower in the 20 mA group (2.5%, n = 94) than in the 35 mA group (6.5%, n = 94) (P < 0.001). While image noise was higher in the 20 mA group (2.1 Hounsfield units versus 1.0 Hounsfield units of the 35 mA-group, P < 0.001), all the anatomic landmarks remained visible in all cases. The over z-axis coverage proportion demonstrated high correlations with the dose proportion for both groups (R = 0.736, P < 0.001 for the 35 mA group and R = 0.660, P < 0.001 for the 20 mA group). The bolus tracking dose-length product proportion demonstrated the strongest positive correlation with the number of monitoring scans (R = 0.93, P < 0.001), while age, diagnostic scan dose-length product, and water-equivalent area showed weak negative correlations (R-values = -0.46~-0.50, P-values < 0.001). Conclusions In low-dose pediatric cardiothoracic CT, the dose proportion of the localizer radiograph can be substantially reduced with a low tube current setting while maintaining image quality. Additionally, minimization of the over z-axis coverage proportion merits attention. The number of monitoring scans is the most significant factor for increasing the radiation dose proportion of bolus tracking, especially in young ages. Graphical Abstract

The role of preoperative cardiac computed tomography (CT) in neonates with pulmonary atresia and intact ventricular septum (PA-IVS) remains unclear. This study was aimed to elaborate the role of preoperative CT-derived anatomical and functional findings in planning treatment strategies in neonates with PA-IVS. The presence of ventriculocoronary arterial connections was evaluated by CT. CT-derived ventricular volumetric parameters were compared and correlated with echocardiographic tricuspid valve (TV) z-score in 12 neonates with PA-IVS. Cardiac CT and echocardiographic findings were compared between definite surgical types (median follow-up, 4 years). Ventriculocoronary arterial connections were identified with CT in 58.3% of cases (7/12) and associated with higher incidence of Fontan procedure (42.9%, 3/7) and high mortality (28.6%, 2/7). The CT-derived and echocardiographic TV z-scores exhibited a high correlation (R = 0.924, p < 0.001). The CT-derived right ventricle (RV) volume and RV-left ventricle volume ratio also displayed high correlations (R = 0.875 and 0.867, respectively; p < 0.001) with echocardiographic TV z-score. More positive echocardiographic TV z-score, high CT-derived RV end-diastolic volume and RV-left ventricle volume ratio, and low CT-derived left ventricular end-diastolic volume were observed in biventricular surgery group (N = 2), compared to Fontan operation (N = 3) and 1.5 ventricular surgery (N = 3) groups, and mortality cases (N = 3). Preoperative CT-derived coronary artery anatomy and ventricular volumetric parameters may supplement treatment planning in neonates with PA-IVS especially when multifactorial decision including echocardiographic TV z-score is in a gray zone.W