Session S35.1

Automated Calculation of Infarct Transmurality

E Heiberg*, H Engblom, M Ugander, H Arheden

Lund University
Lund, Sweden

Infarct size and transmurality are important determinants of prognosis after myocardial infarction. Infarct size can be measured by using contrast delayed enhancement MRI. We have previously developed an automated infarct quantification algorithm that for each voxel assigns an infarct percentage, in contrast to other algorithms, which dichotomously classifies voxels as infarcted or not. Due to partial volume effects one voxel can be partially infarcted. The algorithm was validated in an animal model, computer phantoms and patient images. The aim of this study was to develop a method to calculate infarct transmurality based on a non dichotomous infarct classification, and to compare with the results of manual delineation. To calculate infarct transmurality the infarct percentage was integrated along radial spikes of the myocardium.
Three observers blinded to each others results manually outlined infarct region in 40 patients. A consensus delineation was defined as mean of the three observers. Mean global transmurality for the three observers, and computer algorithm were: 35%, 31%, 38%, and 28%, respectively. One pixel corresponds to around 13% transmurality. Difference compared to consensus delineation were (mean±SD): 0.6±4.4%, -4.2±5.3%, 3.6±5.5%, and -6.7±13%, respectively. The three observers were not statistically different from each other (one-way ANOVA, p=0.18). Transmurality as calculated by the computer algorithm were significantly smaller than the consensus delineation (p<0.05).
Regional analysis were performed in 6 sectors in each MRI slice. For the three observers, and the computer algorithm the mean transmurality in infarcted sectors were: 44%, 31%, 49%, and 30%, respectively. Number of infarcted sectors were: 962, 1324, 789, and 1449. The difference compared to consensus delineation were (mean±SD): 6.7±17%, 2±15%, -9±20%, and 3.2±16%. The three observers differed significantly from each other (one-way ANOVA, p<0.01). The computer algorithm was significantly different from consensus (p<0.01). In conclusion, weighted calculation of transmurality gave smaller global transmurality compared to consensus delineation, but did had the same variability on a regional basis.

(Abstract Control Number: 242)