COMPARISON OF POST-PROCESSING TECHNIQUES FOR OPTIMIZING THE QUALITY OF CARDIAC MAGNETIC RESONANCE T1 MAPPING IMAGES

BackgroundMyocardial T1 mapping is well established as a quantitative, non-invasive tool for myocardial tissue characterization (1). Incidence of abnormal myocardium are common in patients with cardiomyopathies but may additionally stem from anatomical misalignment due to motion (2). The latter influence not truly representative myocardium, thus result falsely elevated values and impair the image quality (IQ) images. Two techniques: MOdified Look-Locker Inversion recovery Motion Correction (MOLLI-MOCO) MOLLI-MOCO deep learning (MOLLI-MOCO-DL), have been proposed target these limitations (3,4).Methods ResultsEighteen were scanned on 3T SIGNA Premier (GE Healthcare, Milwaukee, USA). Three slices acquired short-axis orientation using MOLLI sequence. Images further processed both MOLLI-MOCO-DL. global segmental measured cvi42 (Circle Cardiovascular Imaging Inc, Calgary, Canada) each MOLLI, MOLLI-MOCO, MOLLI-MOCO-DL blinded, cross-sectional analysis. An experienced reader also evaluated IQ 4-point Likert scale where 1 = non-diagnostic IQ, 2 diagnostic many artefacts, 3 few 4 perfect IQ. Normal values, obtained cohort healthy volunteers, used classify segments increased (greater than SD), normal (within SD) or decreased (lower (5). Fleiss Kappa statistic was assess similarity between classification three methods. All 18 successfully scanned, their maps analyzed. example method one subject shown Figure 1. techniques resulted higher compared (Figure 2, A). methods had reasonable agreement score 0.78. classified greater number indicating that they suffer fewer false positives conventional B).ConclusionView Large Image ViewerDownload Hi-res Download (PPT) (3,4). Myocardial Methods B). Eighteen Conclusion