An assessment of the performance of elastography for the investigation of BI-RADS 4 and BI-RADS 5 breast lesions: correlations with pathological anatomy findings.

PURPOSE: To study the diagnostic value of static elastography of the breast (score, histogram, quantitative ratio) in masses classed as BI-RADS 4 and BI-RADS 5 on sonography using the findings from pathological anatomy analyses on the masses as a reference. MATERIALS AND METHODS: A prospective study using a representative sample into 68 masses seen on sonography and their elastography results. For each mass, we determined the elastography colourimetry score (UENO et al. classification, scores 1-3=benign and 4-5=malignant). We studied quantitative elastography parameters based on the Z2/Z1 (fat/lesion) ratio and a histogram showing the pattern of mass stiffness distribution. The results were compared with histology findings (68 lesions assessed, 22 benign lesions and 46 malignant lesions). RESULTS: Elastography was consistent with histology (sensitivity: 73.9%, specificity: 86.4%, PPV: 91.9%, NPV: 61.3%). There were twelve false negatives and three false positives. With the Z2/Z1 ratio, we set a cut-off point of 3.05 for 99.0% specificity. CONCLUSION: Elastography is a reliable technique that is able to assist radiologists in their diagnostic approach to breast pathology.