Characterization of the Physical Image Quality of a Clinical Photon-Counting Computed Tomography Scanner Across Multiple Acquisition and Reconstruction Settings

This phantom study presents a thorough characterization of the physical image quality of a clinical whole-body photon-counting computed tomography (PCCT) scanner. Multiple quality metrics—noise, noise power spectrum (NPS), task transfer function (TTF), and detectability index (d′)—were analyzed across a range of reconstruction algorithms (filtered back projection, FBP, and Quantum Iterative Reconstruction, QIR, with strength levels Q1–Q4), and varying reconstruction kernels (Br40/Br60/Br76/Br98). Both standard (STD, 0.4 mm slice thickness) and high-resolution (HR, 0.2 mm slice thickness) reconstruction modes were assessed. QIR significantly reduced image noise (60–95%) compared to FBP, particularly with sharper kernels. Spatial resolution improved with increasing QIR strength level for smoother kernels and was further enhanced using HR mode with sharp kernels. HR mode exhibited better noise performance than STD with sharper reconstructions, due to the small pixel effect. While STD mode showed higher d′ values for larger objects, HR mode outperformed it for smaller objects and sharper kernels. Compared to a conventional energy-integrating computed tomography system, the PCCT scanner showed superior d′ values under similar settings. Overall, this study highlights the complex interplay between acquisition and reconstruction parameters on image quality, confirms the potential of PCCT technology, and underscores the need for further clinical validation.