Reproducibility of cardiac volumetric parameters derived from fully automatically prescribed image planes: a direct comparison to manual planning at 1.5-T and 3-T MRI.
Authors
Affiliations (6)
Affiliations (6)
- Institute of Diagnostic and Interventional Radiology, Pediatric Radiology and Neuroradiology, Rostock University Medical Center, Rostock, Germany.
- Institute of Diagnostic and Interventional Radiology, Pediatric Radiology and Neuroradiology, Rostock University Medical Center, Rostock, Germany. [email protected].
- GE HealthCare, Munich, Germany.
- GE HealthCare, Barcelona, Spain.
- Department of Radiology, Ludwig-Maximilian University, Munich, Germany.
- University Medical Center, Astana, Kazakhstan.
Abstract
The prescription of cardiac MRI (CMR) image planes is essential for comparable volumetric assessment, but manual planning is time-consuming and error-prone. This prospective single-center study evaluated automated planning and its impact on the reproducibility of volumetric parameters derived from CMR. Fifty-two healthy volunteers (26 males, median age 44.5 years) were divided into a 1.5 T sub-cohort (n = 32, both scans at 1.5 T, interval 2-5 weeks) and a 3 T sub-cohort (n = 20, 1st scan 1.5 T, 2nd scan 3 T, interval 1-2 h). All scans were performed using automated and manual planning with identical protocols, acquiring standard cardiac planes. Subjective quality of plane position was rated blinded by two radiologists. Volumetric analysis was performed fully automatically without corrections on SAX, retrieving right ventricular (RV) and left ventricular (LV) parameters. Wilcoxon matched-pairs signed rank test, intraclass correlation coefficient (ICC), and Bland-Altman analysis were used for statistical assessment. Subjective quality of image planes showed high consistency with good to excellent ratings in both sub-cohorts. Reproducibility of volumetric parameters was good to excellent (all ICC > 0.77) except for LVEF (1.5 T sub-cohort: LVEF manual: 0.323; automated: 0.213; 3 T sub-cohort: LVEF manual: 0.597; automated: 0.742). Overall, reproducibility was better in the 3 T sub-cohort, mainly due to different scan intervals. ICCs were slightly higher compared to manual planning across both sub-cohorts. These trends were also observed in the Bland-Altman analysis. Fully automated plane positioning for CMR provides high-quality image planes, ensuring high reproducibility of cardiac volumetric parameters across both established field strengths. Question The prescription of CMR image planes is essential for a comparable volumetric cardiac analysis, but manual planning is time-consuming and error-prone. Findings Automated plane prescription for CMR provides high-quality image planes, ensuring high reliability and reproducibility of cardiac volumetric parameters across both established field strengths. Clinical relevance Automated plane prescription for CMR reliably provides high-quality image planes, ensuring comparable cardiac volumetric parameters. This technology can simplify the acquisition and promises to reduce variability between follow-up scans, as well as to enhance the availability for patients.