Variable-time positron emission tomography leg protocol to equalize noise for positron emission tomography/computed tomography acquisitions.

Journal Article (Journal Article)

OBJECTIVES: Positron emission tomography protocols conventionally use a constant scan time per bed position (BP). It may be optimal to spend more time scanning some body sections, particularly the more attenuating sections. The relatively consistent tapering of legs may allow a single, variable-time protocol that could reduce or redistribute time and equalize image quality throughout the leg. Reducing the total scan time will limit the opportunity for motion, which can reduce imaging artefacts and improves image fusion, will improve a patient's experience, and will yield the biggest gains in time for the tallest patients. METHODS: Leg dimensions were studied in 55 adult positron emission tomography/computed tomography scans. Image quality versus size was parameterized using bottle phantoms (diameters, 4.5-19.7 cm) filled with identical 18F-fluorodeoxyglucose concentrations, positioned as legs and imaged. An exponential relationship of diameter to noise was fit to the data, which defined the noise-equalizing, variable-time protocol. RESULTS: Of 55 patient leg studies, 94.5% (52/55) had leg diameters that were less than ± 7.5% of the mean leg diameter for other patients of similar height. To equalize noise throughout the leg, relative scan times from superior to inferior as a function of BP are (1, 0.65, 0.45, 0.36, 0.30, and 0.25). CONCLUSION: Variable time compared with a constant-time protocol can require 75% less time for some BPs and half the total acquisition time. A variable-time protocol to reduce time has been implemented.

Full Text

Duke Authors

Cited Authors

  • Wilson, JM; Coleman, RE; Turkington, TG

Published Date

  • September 2011

Published In

Volume / Issue

  • 32 / 9

Start / End Page

  • 868 - 872

PubMed ID

  • 21799371

Electronic International Standard Serial Number (EISSN)

  • 1473-5628

Digital Object Identifier (DOI)

  • 10.1097/MNM.0b013e328347f1a3


  • eng

Conference Location

  • England