Simple protocol tips help make the most of contrast

Minimizing contrast while maximizing quality is a perennial challenge for CT imagers that can be met by following a few simple tips. The key is developing a protocol that makes the most of the contrast being used, said Dr. Michael Lev, director of emergency neuroradiology at Massachusetts General Hospital in Boston, at the Stanford Multidetector-Row CT symposium in San Francisco.

The balance of contrast and image quality becomes particularly important when brain and neck vessels are imaged in cases of potential stroke. The first step is to maximize CT angiography coverage from the aortic arch to the top of the head, Lev said.

"That's to look not only at the blood vessels, stenoses, and circle of Willis but also brain parenchyma," he said. "It turns out that axial source images from CT angiography acquisitions are blood volume-weighted if you give the bolus correctly."

Maximal CT perfusion coverage is also important. In evaluating for stroke, cerebral blood volume and whole-brain source images are used in the same way as diffusion scans, Lev said.

"They basically tell you how much of the brain is already dead," he said. "But cerebral blood flow and mean transit time tell you how much brain is still at risk. What benefit might there be if we could select patient treatment based on the ratio of already dead brain to brain at risk?"

CT perfusion images are notoriously signal-to-noise poor, making it difficult to maximize image quality, he said. A bright first pass is needed. At MGH, Lev prefers 10-mm rather than 5-mm slices for CT perfusion images, and he has seen a 50% boost in SNR as a result.

The advent of 64-slice CT has also spurred a protocol change at MGH for CTA and CTP.

"With 64-slice CT, there are a lot of advantages to doing the CTP first," he said. "There's a big difference in peak signal, and because we already have contrast on board when we perform the CTA, we can often tell if there has been a venous stroke. The delayed CTA blush may also indicate hemorrhage risk."

In addition to thick reconstructions (1.2 to 2.5 mm for image review, rather than the 0.625 mm used for 3D reformats), Lev advised using minimal pitch (always < 1, typically 0.6 to 0.9) and maximum gantry rotation speed during scanning. These techniques, combined with a small focal spot, will help alleviate common problems such as windmill artifact while improving posterior fossa visualization.

Use of high-density contrast (greater than 300 mgI/mL) and a fast injection rate (3.5 to 5.0 cc/sec) will also help boost image quality. A saline push can increase and prolong peak opacity of contrast while decreasing streak artifacts.

"We've tried biphasic injections and bolus shaping, but what really gives us the greatest benefit in these very fast injections and fast scans is the saline push," Lev said.

All imaging is driven by clinical need, and Lev sees that need increasing in neuroradiology, particularly in the case of stroke. Only 3% to 4% of the 750,000-plus strokes annually can be treated by FDA-approved therapy within three hours of onset, he said.

"Where we will see the biggest potential benefit of imaging is visualizing how much of the brain is already dead versus how much of the brain may die in the future," he said. "This mismatch is going to be a very powerful imaging tool, with the potential of giving treatment to hundreds of thousands more patients up to eight or nine hours after stroke onset."