Diagnostic Imaging
October 2003

ADVANCED MR

Use of best imaging options maximizes readability

Sequences affect diagnosis

The practical nature of running a musculoskeletal imaging practice limits the amount of scan time to 30 or at most 60 minutes. There's a constant trade-off, therefore, between the speed of imaging and the appearance of the images.

Musculoskeletal radiologists at Stanford University Medical Center perform knee MR imaging in less than half an hour and usually obtain five acquisitions in different planes with different contrast. They include a sagittal proton density-weighted, fast spin-echo sequence because that's the best option for observing meniscal tears, said Dr. Garry Gold, an assistant professor of radiology at Stanford.

Performing MR imaging of the musculoskeletal system also involves a great deal of variability. Images differ from one institution to another even in the same geographic location, said Dr. David W. Stoller, director of MR and orthopedic imaging at California Pacific Medical Center in San Francisco.

"We've developed more and more sequences and follow routine protocols, but we've deviated from the initial research that was done in many areas of orthopedics. So there are some imaging pitfalls that we have to be careful about. If we haven't used the correct pulse sequences to maximize diagnosis, we may misdiagnose a problem or lose specificity or sensitivity," he said.

Stoller's practice frequently reads scans performed at other institutions that use gradient-echo sequences as if they were sensitive to cartilage or bone marrow edema. While gradient-echo musculoskeletal imaging has its place, fat-suppressed fast spin-echo or short-tau inversion recovery (STIR) are needed for an accurate diagnosis, he said.

The workhorse sequence for orthopedic imaging is fat-suppressed proton density-weighted FSE for T2-weighted contrast imaging. Without fat suppression, FSE for T2-weighting does not reveal bone marrow edema or small cystic fluid collections, perilabral or meniscal cysts, or even cruciate or other ligamentous pathologies, Stoller said. He cautions that fat suppression cannot be added to an FSE T2-weighted sequence with a long TE because signal is replaced by noise.

Although proton density-weighted sequences effectively visualize the articular cartilage, they should not take the place of T1-weighted imaging, according to Stoller. With T1-weighted imaging, radiologists can appreciate the hypointensity of bone marrow edema, neoplasms, and chronic sclerotic changes.

"There are many imaging options, ranging from 2D to 3D, fat suppression, intravenous and intra-articular contrast, and imaging sequences," he said. "We need to be sure we're using the correct ones to maximize diagnosis."