Ventricular contouring of cardiac magnetic resonance imaging is the gold standard for volumetric analysis for repaired tetralogy of Fallot (rTOF), but can be time-consuming and subject to variability. A convolutional neural network (CNN) ventricular contouring algorithm was developed to generate contours for mostly structural normal hearts. We aimed to improve this algorithm for use in rTOF and propose a more comprehensive method of evaluating algorithm performance. We evaluated the performance of a ventricular contouring CNN, that was trained on mostly structurally normal hearts, on rTOF patients. We then created an updated CNN by adding rTOF training cases and evaluated the new algorithm’s performance generating contours for both the left and right ventricles (LV and RV) on new testing data. Algorithm performance was evaluated with spatial metrics (Dice Similarity Coefficient (DSC), Hausdorff distance, and average Hausdorff distance) and volumetric comparisons (e.g., differences in RV volumes). The original Mostly Structurally Normal (MSN) algorithm was better at contouring the LV than the RV in patients with rTOF. After retraining the algorithm, the new MSN + rTOF algorithm showed improvements for LV epicardial and RV endocardial contours on testing data to which it was naïve (N = 30; e.g., DSC 0.883 vs. 0.905 for LV epicardium at end diastole, p < 0.0001) and improvements in RV end-diastolic volumetrics (median %error 8.1 vs 11.4, p = 0.0022). Even with a small number of cases, CNN-based contouring for rTOF can be improved. This work should be extended to other forms of congenital heart disease with more extreme structural abnormalities. Aspects of this work have already been implemented in clinical practice, representing rapid clinical translation. The combined use of both spatial and volumetric comparisons yielded insights into algorithm errors.

Background: Measures of carotid intima media thickness (cIMT) in adults are correlated with adiposity and the metabolic syndrome (MetS) and predict cardiovascular (CV) events. Relations in children are not as well studied. Our objective was to determine the relations of cIMT with body mass index (BMI) and CV risk score in children. Methods: The study included 291 children (158 M/133 F) 6–18 years of age (140 aged 6–11/151 aged 12–18) with measurements of height, weight, waist circumference; fasting lipids, glucose, insulin, and cIMT. A CV risk cluster score was developed from sum of the z-scores of the five MetS components (waist circumference, blood pressure, serum triglycerides, high-density lipoprotein cholesterol, and insulin). Partial Pearson correlation coefficients were adjusted for age, sex, and race. Results: There was no significant age difference in cIMT from 6 to 18 years of age. BMI and CV risk score were significantly correlated (P < 0.0001), and both were correlated with cIMT (r = 0.14, P = 0.02 and r = 0.16, P = 0.006, respectively). Slight age-related differences in associations of cIMT with CV risk score and BMI were explained by unusual values in a few children. Conclusions: These cross-sectional data in normal children show that cIMT was stable from childhood into adolescence. However, both BMI and CV risk score had small, but significant positive correlations with cIMT. Therefore, maintaining normal levels of adiposity and other risk variables may be useful in preventing early changes associated with preclinical atherosclerosis.