Exhibit Hall | Forum 9
Purpose: Processing MOLLI images to measure longitudinal relaxation time (T₁) typically involves pixel-based mapping followed by segmentation of anatomical compartments. This method can be skewed by errant pixels or images, especially in the presence of artifacts. Calculating T₁ from regressions of mean myocardial pixel values significantly increases time and user input but might average out errors stemming from noise and artifacts. This study aims to determine if ease and speed afforded by the mapping method cause significant deviations in measurement compared to the non-mapping method.
Methods: Three 2D short axis (SAX) images were acquired at the basal, mid-ventricular and apical segments of the left-ventricular myocardium in 63 baboon subjects using a MOLLI CMR sequence. Epicardial and endocardial margins were manually drawn on all 11 images in each set. A regression model was formulated from myocardial ROI averages to calculate T₁. This same regression method was used to obtain a T₁ map for each image set. Anatomical regions were segmented using a region-filling algorithm. The two methods were compared using Bland-Altman analysis and Pearson correlation.
Results: Averages of SAX myocardial T₁ values were correlated between the methods (r=0.83, p<0.001). T₁ differences were normally distributed per Shapiro-Wilk test (p=0.16). Bland-Altman analysis revealed that the line of equality was well within the mean difference CI (CI= -8.7 ms to 4.8 ms). Lower and upper limits of agreement were within 5% of the mean value of 1129.8 ms (CI= -54.6 ms to 50.7 ms).
Conclusion: The data demonstrate that there is no difference between the two methods studied. Mapping is often the method of choice because of the time savings afforded by annotating one image compared to eleven. This study shows that the shortcuts in mapping do not significantly affect T₁ measurements compared to the non-mapping method.
Funding Support, Disclosures, and Conflict of Interest: NIA U19AG057758
Image Analysis, Quantitative Imaging, Heart