Premium
Rapid objective measurement of gamma camera resolution using statistical moments
Author(s) -
Hander Trish A.,
Lancaster Jack L.,
Kopp David T.,
Lasher John C.,
Blumhardt Ralph,
Fox Peter T.
Publication year - 1997
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.597928
Subject(s) - imaging phantom , gamma camera , image resolution , optics , pixel , optical transfer function , full width at half maximum , physics , resolution (logic) , bar (unit) , mathematics , nuclear medicine , artificial intelligence , computer science , medicine , meteorology
An easy and rapid method for the measurement of the intrinsic spatial resolution of a gamma camera was developed. The measurement is based on the first and second statistical moments of regions of interest (ROIs) applied to bar phantom images. This leads to an estimate of the modulation transfer function (MTF) and the full‐width‐at‐half‐maximum (FWHM) of a line spread function (LSF). Bar phantom images were acquired using four large field‐of‐view (LFOV) gamma cameras (Scintronix, Picker, Searle, Siemens). The following factors important for routine measurement of gamma camera resolution with this method were tested: ROI placement and shape, phantom orientation, spatial sampling, and procedural consistency. A 0.2% coefficient of variation (CV) between repeat measurements of MTF was observed for a circular ROI. The CVs of less than 2% were observed for measured MTF values for bar orientations ranging from −10° to +10° with respect to the x and y axes of the camera acquisition matrix. A 256×256 matrix (1.6 mm pixel spacing) was judged sufficient for routine measurements, giving an estimate of the FWHM to within 0.1 mm of manufacturer‐specified values (3% difference). Under simulated clinical conditions, the variation in measurements attributable to procedural effects yielded a CV of less than 2% in newer generation cameras. The moments method for determining MTF correlated well with a peak–valley method, with an average difference of 0.03 across the range of spatial frequencies tested (0.11–0.17 line pairs/mm, corresponding to 4.5–3.0 mm bars). When compared with the NEMA method for measuring intrinsic spatial resolution, the moments method was found to be within 4% of the expected FWHM.