[Distinguishing benign and malignant lesions with time-signal intensity curve of dynamic contrast-enhanced breast MRI scanning].

Objective: To determine the diagnostic value of time-signal intensity curve (TIC) in distinguishing breast malignant tumors from benign lesions.
Methods: Forty-four patients with 50 breast lesions were recruited in the study, including 24 pathologically confirmed benign lesions and 26 malignant tumors. All patients received dynamic contrast-enhanced breast MRI scanning a week before surgery. The time-signal intensity curves in the regions of interest (ROI) and eight items of TIC including shape, T peak, E peak, Slope(i), E1, E2, W peak-7, and W peak-9 were compared between benign lesions and malignant tumors. The receive operating characteristic curves (ROC) were depicted for those indicators with significant statistical differences.
Results: Six items of TIC including shape, T peak, Slope(i), E1, E2, and W peak-9 were significantly different (P<0.05) between malignant tumors and benign lesions. The sensitivities of shape, T peak, Slope(i), E1, E2, and W peak-9 for diagnosing breast malignant tumors were 92.3%, 95.83%, 80.77%, 61.53%, 69.23%, and 69.23% respectively. Their specificities were 87.5%, 92.3%, 95.8%, 87.5%, 79.17%, and 79.17% respectively. TIC curve shape, T peak and Slope(i) were better than E1, E2 and W peak-9 in diagnosing breast malignant lesions (P<0.05).
Conclusion: Benign and malignant breast tumors can be differentiated according to TIC. TIC curve shape, T peak and Slope(i) are better than E1, E2 and W peak-9 in distinguishing malignant breast tumors from benign lesions.