To investigate the performance of a deep learning-based algorithm for fully automated quantification of left ventricular (LV) volumes and function in cardiac MRI. We retrospectively analysed MR examinations of 50 patients (74% men, median age 57 years). The most common indications were known or suspected ischemic heart disease, cardiomyopathies or myocarditis. Fully automated analysis of LV volumes and function was performed using a deep learning-based algorithm. The analysis was subsequently corrected by a senior cardiovascular radiologist. Manual volumetric analysis was performed by two radiology trainees. Volumetric results were compared using Bland–Altman statistics and intra-class correlation coefficient. The frequency of clinically relevant differences was analysed using re-classification rates. The fully automated volumetric analysis was completed in a median of 8 s. With expert review and corrections, the analysis required a median of 110 s. Median time required for manual analysis was 3.5 min for a cardiovascular imaging fellow and 9 min for a radiology resident (p < 0.0001 for all comparisons). The correlation between fully automated results and expert-corrected results was very strong with intra-class correlation coefficients of 0.998 for end-diastolic volume, 0.997 for end-systolic volume, 0.899 for stroke volume, 0.972 for ejection fraction and 0.991 for myocardial mass (all p < 0.001). Clinically meaningful differences between fully automated and expert corrected results occurred in 18% of cases, comparable to the rate between the two manual readers (20%). Deep learning-based fully automated analysis of LV volumes and function is feasible, time-efficient and highly accurate. Clinically relevant corrections are required in a minority of cases.

BackgroundMapping of T1 and T2 relaxation times in cardiac MRI is an invaluable tool for the diagnosis and risk stratification of a wide spectrum of cardiac diseases.PurposeTo investigate the global and regional reproducibility of native T1 and T2 mapping and to analyze the influence of demographic factors, physiological parameters, slice position, and myocardial regions on reproducibility.Study TypeProspective single‐center cohort‐study.PopulationFifty healthy volunteers (29 female, 21 male) with a mean age of 39.4 ± 13.7 years.Field strength/SequenceEach volunteer was investigated twice at 1.5 T using a modified look‐locker inversion‐recovery (MOLLI) sequence (T1 mapping) and a T2‐prepared steady‐state free precession (SSFP) sequence (T2 mapping).AssessmentGlobal T1 and T2 values were quantified for the entire left ventricle in three short‐axis slices. Regional T1 and T2 values were measured for each myocardial segment and for myocardial segments grouped by slice position and anatomical region.Statistical TestsTest–retest reproducibility was assessed using intraclass correlation coefficient (ICC) and Bland–Altman statistics. A P value < 0.05 was considered statistically significant.ResultsReproducibility was good for global T1 values (ICC 0.88) and excellent for global T2 values (ICC 0.91). Reproducibility of T1 values was excellent (ICC 0.91) for midventricular slice and good for apical (ICC 0.86) and basal slice (ICC 0.81). Reproducibility of T1 mapping values was highest in the septum (ICC 0.90) compared to the anterior (0.81), lateral (0.86), and inferior (0.86) wall. For T2 mapping, reproducibility was good for all slice positions (ICC 0.86 for midventricular, 0.83 for basal, and 0.80 for apical slice). Reproducibility of T2 mapping was significantly lower for the inferior wall (ICC 0.58) than for septum (0.89), anterior (0.85), and lateral (0.87) wall.Data ConclusionNative T1 and T2 mapping has good to excellent reproducibility with significant regional differences.Evidence Level2Technical EfficacyStage 2