Author: Jennifer Kwak, Chelsea Manica, Justin K. Mikell, Michael Silosky, Wendy Siman 👨🔬
Affiliation: Washington University School of Medicine in St. Louis, University of Colorado Anschutz Medical Campus, School of Medicine, Rocky Vista University 🌍
Purpose:
This study evaluates synthetic planar imaging (synP) from SPECT projections against conventional planar imaging, focusing on detectability, spatial resolution, and feasibility. SynP allows efficient generation of multiple views without increasing acquisition time.
Methods:
Low-contrast detectability and spatial resolution were evaluated for planar and synthetic planar (synP) images using Tc-99m phantoms on a Siemens Pro.specta SPECT/CT system with LEHR collimators.
For low-contrast detectability, planar images and SPECT projections of a NEMA IEC phantom (2 μCi/cc background, two 60-mm spheres; SBR: 4 and 0) were acquired with equivalent durations (1–15 min). SynP images were reconstructed with scatter correction and analyzed visually and via contrast-to-noise ratio (CNR). Background noise was quantified as the relative standard deviation.
Spatial resolution was assessed using a line phantom (5 mCi/cc) imaged at 0–15 cm off-isocenter with 5-min acquisitions. Full-width at half maximum (FWHM) of line profiles was measured for both synP and planar images.
Results:
Visual and CNR-based assessments showed that low-contrast cold sphere detectability in synP was comparable to planar images. For a 5-min acquisition, CNR values were 3.7 in planar and 5.6 in synP.
SynP also had lower noise, with relative noise levels of 3.4% vs. 5% in planar images, but noise was more prominent in the low spatial frequency domain.
Planar images slightly outperformed synP in spatial resolution at 0–15 cm off-isocenter. FWHM values at 0 and 10 cm were 15.6 mm and 9.2 mm (planar) vs. 16.1 mm and 10.5 mm (synP).
Conclusion:
Phantom studies suggest that synP offers higher low-contrast detectability, comparable spatial resolution, lower noise, and shorter acquisition durations compared to conventional planar imaging.
However, higher low-frequency noise in synP images may obscure low-detectability objects in clinical settings. Clinical studies are needed to validate its utility and address these limitations, as phantom results remain promising.