Purpose
Currently, the American Heart Association (AHA) 17-segment model is the preferred clinical method to define and quantify left ventricle (LV) myocardial infarction (MI) size. This method is subjective and can be inaccurate given that segmental approximation assumes a specific percent of infarcted tissue when compared to reference standard post-mortem histopathology.
To improve the accuracy and reproducibility of infarct volume quantification we propose a novel measurement technique based on cardiac MRI images from a porcine model of myocardial infarction. Data were collected from serial MRI exams of Yucatan mini swine over 6 months and endpoint organ harvesting for histopathologic analysis.
Methods
Two observers evaluated four infarct sizing methods: myocardial contouring of post-mortem heart slices, contouring using cardiac MRI, AHA 17-segment model analysis and novel long-axis MRI infarct sizing.
Results
LV infarct sizes ranges were 1.6% - 25.8% (n=10) using reference standard histopathologic infarct sizing. Intraclass correlations (ICC) were calculated between two observers and averaged due to high similarity, ICC > .900. A t-test of .0006 and Bland-Altman plots show statistically significant differences in 17-segment model infarct size compared to histopathologic analysis while no significant difference was found when compared to our new novel method with 0.8198. Linear correlation showed an R 2 of 0.9111 between MRI contoured infarct size and our novel MRI infarct sizing model to predict infarct size as a percentage while the R 2 of the 17-Seg model is 0.8197.
Conclusions
The 17-sgement model provides an inferior quantitative assessment of LV infarct size compared to the proposed long-axis infarct sizing suggesting it maybe a robust and easily implementable quantitative assessment of LV infarct size in advanced imaging.
