How to quantify infarct size on delayed-enhancement MR images: a comparison between visual and quantitative approach

PURPOSE: Our aim was to evaluate the reliability of visual quantification of infarct extent on delayed enhanced magnetic resonance images. MATERIALS AND METHODS: Eighty patients with previous myocardial infarction underwent cine and contrast-enhanced cardiac magnetic resonance imaging. The gadolinium-enhanced images were evaluated using a segmental model with two different methods: a visual score on a 5-point scale (0 no hyperenhancement, 4 hyperenhancement>76% of myocardial wall) and a quantitative analysis based on the manual tracing of infarct contours with automatic threshold analysis. Each segment was also assigned a wall-motion score ranging from 0 (normokinesia) to 4 (dyskinesia). Statistical evaluation was performed. RESULTS: Out of 1,280 segments, 322 (25.1%) showed wall-motion abnormalities with enhancement in 327 (25.5%) evaluated with visual score and in 414 (32.3%) quantitatively. Among segments with normal or mild hypokinesia, 89.2% had a delayed-enhancement scoreor=3. Mean time required for the visual and quantitative approach was 7+/-3 and 18+/-9 min, respectively. There was strong agreement between the visual and quantitative method (k=0.92; p<0.01). CONCLUSIONS: Visual analysis of delayed enhancement is a timesaving approach that is sufficient to assess the transmural extent of infarction. Moreover, it has high correlation with wall-motion abnormalities.