PRIMARY BREAST ABNORMALITIES - SELECTIVE PIXEL SAMPLING ON DYNAMIC GADOLINIUM-ENHANCED MR-IMAGES

PURPOSE: To evaluate a region-of-interest (ROI) analytic method that i nvolves selective sampling of pixels within predetermined ranges of co ntrast material enhancement values (''thresholding'') on magnetic reso nance (MR) images of primary breast abnormalities. MATERIALS AND METHO DS: Dynamic gadolinium-enhanced MR images were obtained in 105 women. ROIs were drawn to outline the full extent of lesions. Relative signal intensity increase was determined on a pixel-by-pixel basis on 1- and 2-minute postcontrast images, as was the maximum relative signal inte nsity increase. Thresholding was used to analyze each ROI, with the up per boundary defined by the highest pixel value and lower boundaries o f 0%-100%. RESULTS: Seventy-one invasive carcinomas and 37 benign:lesi ons were analyzed. Narrower thresholding tie, larger percentage) resul ted in an increase in all enhancement ratios (P < .0005). The enhancem ent ratio on 1-minute postcontrast images differed between benign lesi ons and carcinomas (P < .0005), but there were no significant differen ces:in 2-minute and maximum ratios. Mean enhancement differences betwe en benign and malignant lesions increased with narrower thresholding, but variability also increased linearly. Results of receiver operating characteristic analysis showed that thresholding did not affect the d iagnostic usefulness of enhancement ratios. CONCLUSION: Contrary to cu rrent opinion, selective sampling of the most enhancing areas Of breas t abnormalities may not provide any diagnostic advantage over the use of easily drawn, lesion-encompassing ROIs.