Author/Authors :
Nyathi, Mpumelelo Department of Medical Physics, Faculty of Health Sciences, SMU, Pretoria, South Africa , Sithole, Enoch Department of Physics, Faculty of Pre-Clinical Sciences, SMU, Pretoria, South Africa , Ramafi, Ouma Department of Nuclear Medicine, Dr. George Mukhari Academic Hospital, Pretoria, South Africa
Abstract :
Introduction: The limited resolution of the imaging system causes partial volume effects (PVEs). These results in spreading of
image counts to the neighboring pixels. This phenomenon is called spill-out effect. This study aimed at quantifying PVEs using
ImageJ.
Methods: Technetium-99m solution of concentration of 74 kBq/ml was filled into spheres A, B and C of diameters: 26 mm, 20 mm
and 16 mm respectively. The spheres were imaged mounted inside a Jaszczak phantom filled with activity free water using a
Siemens E-Cam dual head gamma camera. Images were quantified using ImageJ following a two-step method. Step 1: Drawing of
region of interest 1 (ROI 1) closely on the boundary of the planar image to extract images counts before PVEs correction. Step 2:
Drawing region of interest 2 (ROI 2) to extract true sphere image counts. ROI 2 extends from the boundary of ROI 1 by the FWHM
of the imaging system.
Results: The study revealed that PVEs are aggravated by decrease in sphere size. Underestimation of image counts on the 64 × 64
pixels matrix was found to be: 9.7%; 15% and 26% in the order of decreasing sphere size. However, an improvement in the spatial
resolution decreased PVEs (128 × 128 pixels: 6.7%; 12.0% and 22.5%; 256 × 256 pixels: 6.5% ; 9% and 19.3%; 256 × 256 pixels:
6.1%; 8.0% and 18.7% in the order of decreasing sphere size).
Conclusion: ImageJ successfully quantified PVEs attributed to the spill-out effect in planar imaging.
