PET/CT and SPECT/CT fusion imaging: Technical and clinical highlights
Steven Bujenovic, MD, director of nuclear medicine and PET imaging for Our Lady of the Lake Regional Medical Center in Baton Rouge, La., discussed the value of PET during an educational session at the Radiological Society of North America meeting in Chicago.

"Why do we need PET?" he asked. "Looking at physiology has a big advantage over looking at anatomy. Abnormal physiology (behavior) happens before abnormal structure (results)." That's important since 90 percent of FDG PET is for cancer.

"Why not just use PET?" Bujenovic asked -- because physiology can be unpredictable. Sometimes tumors don't take up FDG because they are in a slow growth phase. PET-CT shows muscle uptake and brown fat easily. Radiation therapy protocol (RTP) requires imaging of the target (PET-CT) but also what anatomy to avoid (CT).

Bujenovic showed a series of images from a case where esophageal cancer was found only with PET-CT. "Adding PET and CT makes interpretation of both images better." PET beats CT when it comes to spatial resolution. PET also provides high contrast resolution, but CT shows high resolution for shades of grey. "The Holy Grail is an image one can use to modify therapy." That includes anatomy, and characteristics of the tumor, such as DNA metabolism and amino acid metabolism.

With PET-CT fusion, there are several potential obstacles. You must be careful about patient movement. Table movement from worn gear can occur in the Z axis with resulting shifts of PET and CT data. Another concern is table deflection. A twist between CT and PET scanners can adversely affect the X and Y axis alignment. And, your field service engineer might not correctly set up the scanner software.

Most scanner software does not provider visualization of fusion, so Bujenovic recommends that you consider using a phantom that can register both PET and CT.

Proper positioning methods are similar to those for RTP, such as three point laser, fiducial markers, immobilization devices, and a flat table insert. Most RTP software won't accept fused data set. But both are important because "PET finds and CT defines."

Some physicians say they don't need fusion software pr PET-CT. There are several PET-CT artifacts to consider. Physiologically, there is saliva, urine, and gastric uptake. On the technical side, there is misregistration, and beam hardening.

A fusion image is also darker with a loss of resolution due to the larger pixel size. Cardiac uptake is variable, Bujenovic said. For example, the left ventricle is seen after four hours 80 percent of the time which is important to know for cardiac viability scans. You're always going to have some scatter, he said, but to establish scatter, you must know the source of strength and surrounding attenuation. There is more scatter with larger patients.

For successful interpretation, all clinical signal algorithms should be described and used to their utmost diagnostic capabilities with as low a radiation dose as possible. "Reading shades of grey isn't doing the patient a benefit," he said. "Interpret a fusion image as an integrated report but the interpreter should be familiar with the technology's strengths and weaknesses."