Influence of myocardial region of interest definition on quantitative analysis of planar 123I-mIBG images.

ABSTRACT

PURPOSE: In planar (123)I-mIBG myocardial imaging, definition of the heart region of interest (ROI) is a critical step in quantifying uptake. The present study evaluated the impact of changes in heart ROI size on quantitative results in subjects with good and poor uptake. METHODS: Reference irregular whole-heart and square upper mediastinum ROIs were defined visually on 531 planar (123)I-mIBG images. Based on the reference heart ROI, an automated program created two other ROIs one larger (+1 pixel) and one smaller (-1 pixel), the stated numbers representing the spacing intervals between each epicardial boundary pixel. Two additional smaller ROIs (-2 and -3 pixels) were drawn for the 100 images (19 %) with a heart/mediastinum (H/M) ratio ≤1.30. The number of pixels, the counts per pixel, and the H/M ratio for each heart ROI were calculated and compared to that in the reference ROI. Washout rate and changes as a function of ROI size were also calculated for 110 subjects who had both early and late images. RESULTS: The mean changes in heart ROI size ranged from +19.0 % for the +1 pixel to -44.4 % for the -3 pixels ROI. For the +1 and -1 pixel ROIs, mean relative counts per pixel changes were -1.2 % and +0.7 %, respectively, with corresponding ranges of change in the H/M ratio of -0.12 to +0.05 and -0.05 to +0.11. For scans with H/M ratio ≤1.30, mean relative counts per pixel changes were 0, 0, -0.7 %, and -1.4 % for the four ROIs, with range of change in the H/M ratio from -0.13 to +0.05. Mean washout rates were almost identical for the reference ROIs (45.0 %) and the +1 pixel and -1 pixel heart ROIs (44.9 % and 45.1 %, respectively). CONCLUSION: Significant changes in the size of the best visually defined heart ROI produce minimal, clinically inconsequential changes in the H/M ratio and washout rate, even in subjects with significantly reduced myocardial uptake of (123)I-mIBG.