Musculoskeletal specialists push spatial resolution to limit

April 15, 2005

Opinions differ on how much the signal-to-noise ratio actually improves at 3T MR imaging. Body imagers may not obtain the theoretical doubling they had hoped for, because of inhomogeneities across the field-of-view. Nonetheless, many musculoskeletal radiologists are finding that the improved signal from 3T generates substantially clearer images (Figure 1), particularly on small-FOV scans. They also claim that these highly detailed views are likely to improve patient management.

Opinions differ on how much the signal-to-noise ratio actually improves at 3T MR imaging. Body imagers may not obtain the theoretical doubling they had hoped for, because of inhomogeneities across the field-of-view. Nonetheless, many musculoskeletal radiologists are finding that the improved signal from 3T generates substantially clearer images (Figure 1), particularly on small-FOV scans. They also claim that these highly detailed views are likely to improve patient management.

Dr. Kent Sanders, an assistant professor of radiology at the University of Utah Health Sciences Center in Salt Lake City, began using a 3T scanner for his clinical work about 18 months ago. With no previous experience of 3T and little published literature on high-field musculoskeletal imaging to refer to, Sanders considered the transition from 1.5T to 3T a step into the unknown. Uncertainty about the value of 3T in his clinic was compounded by his colleagues' gloomy forecasts about the technical difficulties ahead.

"A lot of people told me, 'If you do 3T MRI in the knee, the chemical shift artifact is going to be so bad that it will obscure a great deal of the anatomy.' But none of the predictions about artifacts have come to fruition, and I don't have any problems with chemical shift obscuring structures in joints," he said. "So I was quite pleasantly surprised at the improvement in spatial resolution and the anatomical definition of what we are seeing."

Sanders uses the improved SNR from 3T to attain extremely high resolution images of small joints (Figure 2). His small-FOV acquisitions have far higher spatial resolution than would be possible at lower field strengths, he said, and they demonstrate the real benefit of 3T for musculoskeletal work.

"It's just amazing compared to small-FOV imaging at 1.5T, where the images are not nearly as crisp, and you cannot resolve the structures. Subtle differences, particularly articular cartilage, show up so much better on 3T. It has made a dramatic difference to us looking at small parts."

The additional detail seen on 3T MR has improved radiologists' diagnostic accuracy, especially in ligament injuries of the wrist, fingers, and thumb. The high-field MR scanner also helps confirm or rule out clinicians' suspicions of small-joint pathology. If, for example, a podiatrist requested an MR scan to look for sesamoiditis in a patient's toe, the 3T images would be near-histologic in quality, Sanders said.

The clarity of musculoskeletal images is so good that Sanders is eager to expand his repertoire. For now, however, he remains limited by the lack of dedicated coils for joint imaging. Three-T imaging has been the province of neuroradiologists for so long that scanner manufacturers are struggling to meet the needs of body imagers.

"I would cut off my left foot if it meant I could get a shoulder coil out here. Shoulder imaging is such a large part of what we do in musculoskeletal imaging," he said.

Dr. Thomas Link, an associate professor of musculoskeletal radiology at the University of California, San Francisco, also looks forward to wider availability of 3T shoulder coils. Three-T MRI of the shoulder could be particularly useful for evaluating internal joint structure, he said. Small tears to the shoulder labrum, for instance, are often overlooked on 1.5T MRI. Studies of the hip performed at 3T at UCSF revealed the labrum in

far greater detail, suggesting that the same would be true in the shoulder.

Link has been performing 3T musculoskeletal MRI at UCSF since 2003 and uses it routinely in clinical practice for imaging chronic knee pain and sports-related injuries. A 3T MR scan can examine early cartilage damage and improve assessment of ligamentous and meniscal pathology (Figures 3 and 4). The high-field system has also been used for some spine and wrist examinations.

The 3T images are more detailed than those obtained at 1.5T, Link said. But musculoskeletal radiologists switching to higher field strength scanning should not expect dramatic improvements from day one. In addition to using dedicated coils, they must tailor their imaging sequences to suit the 3T hardware and software.

"In the beginning, I was really disappointed about image quality, to be honest," Link said. "Then we worked on the sequences, optimized them, and the images started to look quite a lot better. It's not as simple as taking the sequences you use at 1.5T and just applying them straightaway."

Prof. Dr. Siegfried Trattnig, medical director of the MR Center of Excellence at the Medical University of Vienna, agrees that time spent honing sequences for 3T applications is well spent. In late 2004, Trattnig helped colleagues devise appropriate protocols for joint examinations.

"This process of optimization can take two to three weeks, but it is worth doing because you can improve your protocols and image quality significantly," he said.

Trattnig has been conducting research on 3T musculoskeletal MR for four years, with particular emphasis on high-resolution imaging of cartilage. Resolution of cartilage layers in joints smaller than the knee is generally difficult at 1.5T. Using a gradient insert that boosts gradient strength to 200 mT/m, however, Trattnig and colleagues can examine cartilage around the small interphalangeal joints of the finger.

"Normally, the standard gradient strength is around 30 mT/m. This high gradient strength really allows you to perform MR microscopy on these joints," he said.

This procedure could be used to assess early signs of arthritic disease in small joints of patients with rheumatoid arthritis, he said. Such detailed images might also reveal how cartilage reacts to the onset of inflammatory disease and what happens to its structure during treatment.

Trattnig's group uses a special sodium coil to image cartilage at 3T. The technique can visualize proteoglycan concentration, offering a biochemical assessment of the cartilage layers. It provides a viable alternative to a contrast-enhanced MR method known as dGEMRIC imaging.

"This is a new field that is very promising," he said. "It is a very demanding technique, and you need a sufficient SNR to get good image quality because sodium ions are very rare in comparison to protons. This already works in specimens, and we are now ready to move from in vitro studies to in vivo applications in volunteers."

While most 3T musculoskeletal MRI at the Medical University of Vienna has been done for research purposes, the benefits to clinical practice are already apparent. The traumatic surgery and orthopedic departments regularly perform cartilage repair procedures. While scanning patients from both departments for follow-up, Trattnig discovered that the 3T images revealed fine fissures in the newly implanted cartilage that were not previously seen on standard MR. He introduced a new scoring system, based on high-resolution imaging for follow-up monitoring, which covers how the implant has filled the defect, its integration with adjacent native cartilage, its structure, and its surface appearance.

"In the beginning we did this purely as research because we wanted to look at what was possible with 3T and what we might develop as a new technique. But when we realized that 3T MRI was so much better than 1.5T MRI for follow-up, we changed the procedure. Now patients with cartilage implants are scanned routinely using 3T," he said.

Patient management strategies may also change at the University of Utah when Sanders gets his 3T shoulder coil. He predicts that the resolution of shoulder MRI at 3T could be so high that it might eliminate the need for an arthrography examination, reducing the invasiveness of patient studies and freeing time in the fluoroscopy schedule. The quality of small-joint MRI may already be sufficient to abandon arthrography in certain applications such as internal derangement of the wrist.

"If I do this on a 1.5T scanner, I usually also do an arthrographic study to improve the specificity and resolution of the things that I find," he said. "But my feeling is that you probably don't need to do arthrography on the 3T system. I can see the tangential zone on the articular cartilage quite distinctly, so I don't really need contrast to outline the surface."

Results from 3T MR may also influence clinical practice outside the radiology department. If patients with osteochondral damage undergo a high-field MR scan, for example, orthopedic surgeons can check the true extent of any cartilage damage and tailor their surgery accordingly, Link said.

Debate over the true value of images with exquisite spatial resolution is likely to continue. But for patients whose livelihood might depend on a surgeon's accuracy with a scalpel, 3T clinical musculoskeletal imaging is entirely justified, Sanders said.

"The people who go to see a hand surgeon because they have problems with their middle finger generally have a compelling reason to get the most thorough workup possible. It's usually going to be a professional golfer, basketball player, musician, or elite rock climber, someone who has a huge investment in having that finger work perfectly," he said.