Optimal uniformity index selection and acquisition counts for daily gamma camera quality control

Murray, Anthony W.; Barnfield, Mark C. and Thorley, Penelope J. (2014) Optimal uniformity index selection and acquisition counts for daily gamma camera quality control.

View this record at http://insight.cumbria.ac.uk/1667/
Official URL: 10.1097/MNM.0000000000000167

Abstract

Introduction: The purpose of this study was to investigate the optimized use of common uniformity indices [National Electrical Manufacturers’ Association (NEMA) indices (differential and integral), Cox–Diffey and the coefficient of variation (CoV)]. Methods: The indices were calculated for induced [localized two-dimensional (2D) Gaussian and gradient] artefacts added to three image sets (5, 10 and 15 million counts), each containing 25 extrinsic images, using Matlab. The intensity of the induced artefacts was varied between a 1 and 10% drop in pixel counts. The induced artefacts simulated photomultiplier tube [10 cm full width at half maximum (FWHM)], smaller focused artefacts (2.5 cm FWHM) and gradients artefacts. Results: For five million count acquisitions, the Cox–Diffey, CoV and NEMA integral indices detected the 6% 2D Gaussian artefacts [10 cm full-width at half-maximum (FWHM)], whereas the NEMA differential index performed relatively poorly. NEMA differential and integral indices performed equally well at detecting smaller 2D Guassian (2.5 cm FWHM) artefacts. The 10% artefact was the minimum artefact detected by both indices for five million count acquisitions. The Cox–Diffey and CoV indices did not detect any artefacts for five million acquired counts. The CoV index performed best at detecting gradient artefacts at five million acquired counts. Conclusion: This work provides evidence that daily quality control can be acquired with as few as five million counts while maintaining the same ability to detect both chronic and acute nonuniformities compared with higher count acquisitions. A combination of the NEMA integral and the CoV indices gives the optimal selection of uniformity indices for detecting a range of artefact forms and intensities.

Item Type: Article
Keywords: 001 Knowledge, 616 Diseases & treatment (incl. counselling, radiography & medical imaging, critical & palliative care)
Members: University of Cumbria
Depositing User: ULCC Admin
Date Deposited: 08 Nov 2016 13:05
Last Modified: 08 Nov 2016 13:05
URI: http://collections.crest.ac.uk/id/eprint/12378

Actions (login required)

Edit Item Edit Item