Researchers have demonstrated the first calibration system for positron emission tomography (PET) scanners directly tied to national measurement standards.
According to the researchers from the National Institute of Standards and Technology (NIST), better calibrations of the machines can increase the accuracy of their diagnostic images by several times.
AdvertisementNIST's technique, created over the past few years, is used to calibrate devices, 'phantoms,' specifically for PET scanners.
The system can be used fine-tune PET scanners that find cancers and track the progress of treatments, among other diagnostic applications. It can also be used to ensure the accuracy of some of the newest scanners on the market.
Phantoms are simply blocks of materials known to respond to X-rays in a consistent, known manner that's similar to the way human tissues respond.
They are commonly used to examine medical imaging devices such as X-ray scanners.
PET phantoms are considered to be more complicated because the scanners work by detecting radioactive materials injected in the patient.
The PET phantoms are lightweight hollow cylinders. They contain a small amount of radioactive germanium, which glows brightly in a PET scanner's readout. Measuring its brightness each day provides a benchmark for scanner sensitivity that allows medical technicians to compare patient scans taken on different days more effectively.
Brian Zimmerman, part of the NIST group, says that the team is working on calibrating two more commercial phantom designs for Sanders that will be used with other types of clinical PET scanners. Also, the increased availability of NIST measurement traceability for clinical use could help save time and money for patients, he adds.
"Doctors are realizing putting effort into this kind of quantitative work can reveal differences of 5 percent in tumor size, rather than the 20 percent they look for now. It might help them to make decisions more quickly concerning whether a treatment or a clinical trial is working out," said Zimmerman.