ASSESS SPECT SYSTEM'S PIXEL-WISE TOMOGRAPHIC RESOLUTION THROUGH ANALYSIS OF PROJECTION PROBABILITY DENSITY FUNCTION

Multi-pinhole (MPH) SPECT systems are widely used to improve detection sensitivity while maintaining high spatial resolution. However, image resolution in such systems is commonly characterized using single-point metrics, such as the point spread function (PSF) full width at half maximum (FWHM) estimated at the center of the field of view (FOV). While useful, this single-point estimation fails to capture the spatially variant and tomographic aspect of resolution. The resolution of a SPECT system depends critically on having sufficient angular sampling for each point within the FOV. Incomplete angular sampling can lead to artifacts, distortions, and poor resolution, even when individual projection views are sharp or of high resolution. This issue is particularly pronounced in multi-pinhole systems where the projection of different regions of the FOV onto the detector can be highly non-uniform. Therefore, a metric that can quantify the angular sampling completeness at image pixel level is needed for a more accurate assessment of tomographic resolution. In this work, we propose a novel, spatially variant metric, the Angular Sampling Completeness Index (ASCI), which provides a pixel-by-pixel measure of tomographic sampling quality derived from the projection probability density functions (PPDFs). When interpreted jointly with PPDF beam width (quantified by FWHM), ASCI enables a more comprehensive assessment of spatial resolution performance in MPH-SPECT systems. We validate the proposed framework by evaluation an analytically modelled two-dimensional MPH SPECT configuration with two pinhole diameters under both stationary and rotational detector acquisition schemes. By relating ASCI maps computed over selected PPDF FWHM ranges to the distinguishable hot-rod sizes in reconstructed images, we demonstrate a strong correspondence between angular sampling completeness, intrinsic beam width, and reconstructed image resolution. These results indicate that ASCI serves as an effective predictor of spatially varying image quality in MPH-SPECT.