Pediatric Cardiac Magnetic Resonance Reference Values for Biventricular Volumes Derived From Different Contouring Techniques

Abstract

Background Measurement of ventricular volumes and function using MRI is an important tool in pediatric congenital heart disease. However, normal values for children are sparce and analysis methods are inconsistent. Purpose To propose biventricular reference values in children for two MRI postprocessing (contouring) techniques. Study Type Retrospective. Subjects A total of 154 children from two institutions (13.9 ± 2.8 years; 101 male) that were referred for a clinical MRI study. Field Strength/Sequence 1.5 T; balanced steady‐state free precession (bSSFP) sequence. Assessment Left ventricular (LV) and right ventricular (RV) end‐diastolic and end‐systolic volumes (LVEDV, LVESV, RVEDV, RVESV) and end‐diastolic and end‐systolic myocardial mass (LVEDMM, LVESMM, RVEDMM, RVESMM) were measured from short‐axis images using two contouring techniques: 1) papillary muscles, trabeculations and the moderator band were included in the ventricular blood volume and excluded from the myocardial mass, 2) papillary muscles, trabeculations and the moderator band were excluded from the ventricular volume and included in the ventricular mass. Statistical Tests Univariable and multivariable linear regression models were used to evaluate relationships between sex, weight, height, body surface area (BSA) and age and volumetric results. Reference graphs and tables were created with the LMS‐method. Contouring techniques were compared by intraclass correlation, regression analysis and Bland–Altman plots. A P value < 0.05 was considered statistically significant. Results Height and BSA were significantly associated with LVESV (method 1) and with LVEDV and RVEDV (method 2). LVESV (method 2), RVESV (both methods), RVEDV (method 1), and LVEDMM and RVEDMM (both methods), showed significant associations with height and weight. LVSV and RVSV (both methods) were significantly associated with BSA and weight. RVESV (method 1) was significantly associated with age. Gender showed significant associations for all parameters. Data Conclusion The proposed pediatric reference values can be used in the diagnosis and follow‐up of congenital or acquired heart disease and for research purposes. Evidence Level 3 Technical Efficacy Stage 2

Cardiovascular magnetic resonance reference values of right ventricular volumetric variables in patients with hypoplastic left heart syndrome

“…Regression results from this study show that height is strongly associated with RV volumes compared to other anthropometric or demographic parameters. Similar results were shown in a recently published study from our group in a cohort of healthy children and adolescents that showed that particularly height but also weight showed strong associations with ventricular volumes [12] . Furthermore, It is known that extreme weight (obesity or underweight) can affect CMR results [19] .…”

“…Height or the calculation of the ideal BSA might therefore be more appropriate for indexing of CMR-derived volumetric measurements [19] . Taking into account these consideration, our results [12] , and research from other groups [19] , [21] , it was decided to create reference curves and tables with respect to body height of HLHS patients. RV mass, however, was most strongly associated with patient age suggesting that time might be more important for RV myocardial changes.…”

“…To assess reproducibility, the intraclass correlation coefficient (ICC) was calculated. The Lambda-Mu-Sigma (LMS)-method of Cole and Green was applied to create centile graphs and tables for RVEDV, RVESV, RVMM, and RVEF as described in previous studies [11] , [12] . An extended version of this method is implemented in the R package GAMLSS which was used for the analysis [13] .…”

Editorial for “Pediatric Cardiac Magnetic Resonance Reference Values for Biventricular Volumes Derived From Different Contouring Techniques”

“…In the current issue of JMRI, Voges et al 8 present cardiac biventricular reference values for the above discussed, two different segmentation methods, in a study of 154 healthy children who had undergone 1.5T cardiac MRI. The authors analyzed short‐axis cine images and derived EDV, ESV, ejection fraction, and end‐systolic and end‐diastolic myocardial mass for both LV and RV.…”

Editorial for “Multivendor Evaluation of Automated MRI Postprocessing of Biventricular Size and Function for Children With and Without Congenital Heart Defects”

“…Ejection fraction was relatively normal in these subjects but might affect results. Age, BSA, contouring technique (trabeculae or not), and gender may affect ventricular evaluation, 9 and further exploration of those factors would also be welcome.…”