To assess the usefulness of STIR (short tau inversion recovery) imaging in breast MRI (magnetic resonance imaging).

We retrospectively reviewed T1- and T2-weighted (T1WI, T2WI), STIR, and dynamically enhanced images of 44 pathologically confirmed breast lesions (benign, 13; malignant, 31) in 36 patients. We selected the dynamically image which best depicted a particular lesion, and then made hard copy of the corresponding T1WI, T2WI, and STIR images. Using the dynamically enhanced image as a standard, we analysed these in terms of parenchymal pattern, lesion detectability, differentiation between benign and malignant lesions, extent, multifocality, and the ductal system. The results were statistically analyzed.

In 33 of 44 cases (75%), detectability was greater at STIR imaging than at T1- and T2WI, especially in fibrofatty or fatty breast (14/14 cases, p<0.05). STIR images did not always differentiate between benign and malignant lesions, and extent (50%) and multifocality (46%) were commonly exaggerated compared with T1- and T2WI. In 18 of 44 cases (41%), STIR images suggested the presence of ductal structures.

For the detection of lesions, STIR imaging was more useful than T1- and T2WI, though STIR did not differentiate between benign and malignant lesions. The extent and multifocality of a lesion were exaggerated on STIR images, compared with T1- and T2WI.