[Multi-variated analysis of differential diagnosis in ultrasonography of idiopathic granulomatous mastitis and invasive ductal carcinoma].

Objective: To evaluate the differential diagnosis of idiopathic granulomatous mastitis (IGM) and invasive ductal carcinoma. Methods: The ultrasonographic data of 37 IGM patients and 50 cases of IDC were analyzed retrospectively. The shape, growth direction, margin, internal echo, posterior echo, calcification, Adler blood flow classification, PSV(peak sestolic velocity), RI (resistance index)and elasticity scores were analyzed by χ (2) test and independent sample t test. The optimal cutoff values of age, PSV and RI were calculated by receiver operating characteristic (ROC) curve. Logistic regression analysis was used to calculate the odds ratio ( OR ) of ultrasonic variates in the diagnosis of both diseases. Results: There were no significant differences in the shape, margin, internal echo and blood flow grading between the two groups. The age, lesion growth direction, posterior echo, calcification, PSV, RI and elasticity were statistically different. The cut-of value of Age, PSV and RI were 38.5 years old, 13.20 cm/s, and 0.655. Logistic regression multi-variated analysis revealed that elastic score ( OR =9.806) had the best value of the differential diagnosis, as well as calcification ( OR =6.937), posterior echo decay ( OR =4.613), RI ( OR =3.257), lesion growth orientation ( OR =3.198), and PSV ( OR =1.202). Lesion shape, margin, internal echo, and Adler blood flow classification did not help in differential diagnosis. Conclusion: Ultrasound multi-parameter analysis has high value in IGM and IDC differential diagnosis.