The value of multimodal MRI in differential diagnosis of pure fibroadenosis and mixed fibroadenosis in the breast / 中华放射学杂志

ABSTRACT

Objective To explore the value of multi-modal MRI in the differential diagnosis of pure fibroadenosis and mixed fibroadenosis in the breast. Methods Forty female patients who underwent 3.0 T MRI within 1 week before sugery and confirmed as breast fibroadenosis by pathology in the General Hospital of the PLA Rocket Force from January 2014 to May 2016 were retrospectively analyzed in this study. There were 20 cases of pure fibroadenosis which including mass type and non-tumor type, 10 cases per type. Twenty cases of mixed fibrous adenosis which including 4 cases of mass type and 16 cases of non-mass type. According to the breast imaging reporting and data system-MRI standard, conventional MRI features, time intensity curve (TIC) types and ADC values of the lesions were observed. MRI features and ages of pure fibroadenosis and mixed fibroadenosis were compared using χ2 test (qualitative data) and independent sample t test (quantitative data), P<0.05 was considered statistically significant. Statistically significant parameters were then used to perform logistic regression analysis to evaluate predictive value . The efficacy of each MRI parameter in the differential diagnosis of pure fibroadenosis and mixed fibroadenosis was analyzed by ROC. Results Statistically significant differencein the size(P<0.05) but no differences in the shape, T2WI manifestation, marginal, internal enhancement, early enhancement curve, and late enhancement (P>0.05) were observed between pure fibroadenosis and mixed fibroadenosis. There was no significant differences in distribution, internal enhancement, early enhancement curve and late enhancement curve between non-tumor type pure fibroadenosis and mixed fibroadenosis (P>0.05). There were significant differences in age, ADC value and peak signal intensity(P<0.05) while no significant differences in early enhancement rate, maximum enhancement rate and peak time (P>0.05) between patients with pure fibroadenosis and mixed fibroadenosis. Logistic regression analysis suggested that the peak signal intensity was closely related to age. It revealed a positive correlation between ADC value, peak signal intensity and the possibility of mixed fibroadenosis. The regression coefficient value, Wals value, and P value of the ADC value were 3.652, 4.363 and 0.034, respectively. The regression coefficient value, Wals value, and P value of the peak signal intensity were 0.005, 5.463 and 0.019, respectively. The area under ROC curve of ADC value, peak signal intensity, ADC value combined with peak signal intensity were 0.697, 0.701 and 0.786, respectively. Conclusions Significantly differences of peak signal intensity and ADC value were observed in mixed fibroadenosis compared with pure fibroadenosis. The combination of ADC value and peak signal intensity had the highest efficacy in predicting pure and mixed fibroadenosis.