Objective:

To investigate the different imaging features of mammography with suspicious calcification of fibroadenoma and breast cancer

and to explore its value in the identification of fibroadenoma and breast cancer.

Methods:

Patients with suspicious calcification during preoperative mammography from Jun. 2014 to Sep. 2016 with biopsy or surgical pathology confirmed fibroadenomas and

breast cancer were retrospectively included. Using univariate analysis to investigate the relationship between image characteristics and pathology results

screen out meaningful imaging features to perform logistic regression analysis and to build prediction model

and draw receiver operating characteristic (ROC) curve to evaluate the effectiveness of image characteristics.

Results:

The proportion of calcification and masses in breast cancer group was higher than that in fibroid adenoma group

and the features of unclear mass boundaries were more obvious. In addition

polymorphic calcification was significantly more in breast cancer group than in fibroid adenoma group. Logistic regression analysis showed that the clear-unclear mass boundary was a significant predictor for breast cancer (OR=2.84

95% CI: 1.35-5.97

P

=0.006). The ROC curve was established based on the predicted probability of the logistic regression model. The area under curve (AUC) was 0.7420.073 (

P

=0.005

95% CI: 0.599- 0.885).

Conclusion:

For suspected calcification of mammography

the appearance of mass

the boundary features of the combined mass

and whether there are small polymorphic calcifications are of high value for the differential diagnostis of fibroadenoma calcification and breast cancer.