Objective To investigate the diagnostic value of superb microvascular imaging(SMI), advanced dynamic flow(ADF) and color Doppler flow imaging(CDFI) in breast microcarcinoma, and to compare the correlation coefficients between these three indicators and postoperative pathological tumor microvascular density(MVD). Methods A total of 85 patients(87 lesions) with BI-RADS 4 and the maximum diameter≤1 cm were selected, of which, 46 lesions were benign(benign group) and 41 lesions were malignant(malignant group). All the patients and the corresponding lesions were examined by SMI, ADF and CDFI. Blood flow grading was performed on the images using Adler grade of blood flow, and the difference of blood flow among the 3 methods was compared. The expression level of MVD in pathological tumor tissues was detected and analyzed for its correlation with Adler classification by three detection techniques. Results The areas under the ROC curve(AUR) of CDFI, ADF and SMI were 0.694, 0.705 and 0.776 respectively based on the gold standard with pathological diagnosis. The sensitivity, specificity, positive predictive value, negative predictive value, and accuracy of CDFI were 78.0%, 54.3%, 60.4%, 73.5%, and 65.5%, respectively; those of ADF were 75.6%, 60.9%, 63.3%, 73.7% and 67.8%, respectively; and those of SMI were 78.0%, 69.6%, 69.6%, 78.0% and 73.6%, respectively. The Adler grades of CDFI, ADF and SMI were positively correlated with MVD(P<0.05), which the highest correlation coefficient between SMI and MVD (r=0.430, P<0.001). Conclusions SMI is superior to ADF and CDFI in detecting the abundance of breast microcarcinoma, and has the highest correlation coefficient among those 3 detection techniques with tumor pathological MVD, which indicates that SMI may be used for differential diagnosis of breast microcarcinoma and indirectly evaluate the prognosis of patients. Key words: Superb microvascular imaging; Advanced dynamic flow; Color Doppler flow imaging; Breast microcarcinoma; Adler grade of blood flow; Microvessel density
Read full abstract