Diagnostic value of breast specific gamma imaging with combined visual and semi-quantitative method in patients with breast carcinoma

Abstract

Objective To evaluate the value of breast specific gamma imaging (BSGI) in detecting primary breast cancer by combined visual and semi-quantitative method. Methods A total of 106 patients suspected of having breast cancer underwent BSGI. The diagnosis was confirmed by histological examination. Interpretation for BSGI was based on the visual (grades 1-5) and semi-quantitative assessment. For patients assessed as grade=3 or above, the ROI was drawn on craniocaudal and mediolateral oblique images respectively, and the count ratio of lesion to adjacent normal breast tissue (L/N) was calculated. The higher value of L/N was selected as index quantitative parameter. The optimal visual grade and L/N cut-off value were obtained through ROC analysis. The sensitivity and specificity of BSGI based on visual-only and combined visual and semi-quantitative assessment were evaluated respectively. Two-sample t test was used to analyze the data. Results Postoperative pathological results confirmed 34/106 cases (49 benign lesions) as benign and 72/106 as breast cancer (73 malignant lesions and 4 benign lesions). The optimal visual grade was 3.5, with ROC AUC of 0.901 (95% CI: 0.838-0.965), sensitivity of 94.4% (68/72) and specificity of 73.5% (25/34). The cut-off value of L/N was 1.92, with ROC AUC of 0.870 (95% CI: 0.756-0.984). The L/N value of patients with breast cancer was significantly higher than that of benign cases (3.29±1.30 vs 2.02±0.65; t=5.63, P<0.05). The sensitivity and specificity of combined visual and semi-quantitative method were 91.7%(66/72) and 88.2%(30/34) for diagnosing primary breast cancer. Conclusion Combined visual and semi-quantitative method on BSGI might improve the specificity for the diagnosis of breast carcinoma. Key words: Breast neoplasms; Radionuclide imaging; Mammography