Image quality and dose reduction with photon counting detector CT: Comparison between ultra-high resolution mode and...

The purpose of this study was to assess the image quality and dose reduction potential of ultra-high resolution (UHR) mode compared with standard mode, both available on a commercial photon-counting detector computed tomography (PCCT) scanner.

Images were acquired on a PCCT with a phantom using UHR and standard modes at three dose levels (3/6/12 mGy). Raw data were reconstructed using soft tissue (Br36) and bone (Br68) reconstruction kernels and 0.4-mm slice thickness. Noise power spectrum (NPS) and task-based transfer function (TTF) were calculated to assess noise magnitude, noise texture (f_av), and spatial resolution (f₅₀), respectively. Detectability indexes (d') were calculated to model the detection of two abdominal lesions for a Br36 soft tissue reconstruction kernel and three bone lesions for a Br68 bone reconstruction kernel.

At all dose levels, noise magnitude values were lower with UHR than with standard mode (mean difference, -18.0 ± 2.6 [standard deviation (SD)] % for Br36 and -33.9 ± 2.3 [SD] % for Br68). Noise texture was lower with UHR than with standard mode (mean difference, -4.2 ± 0.9 [SD] % for Br36 and -16.0 ± 1.8 [SD] % for Br68). For the solid water insert and Br36, f₅₀ values were similar for both UHR (0.34 ± [SD] 0.04 mm^-1) and standard (0.33 ± [SD] 0.04 mm^-1) modes. For Br68, f₅₀ values were greater with UHR than with standard for iodine (mean difference, 18.5 ± 1.9 [SD] %) and bone (11.7 ± 5.7 [SD] %) inserts. For all simulated lesions, d' values were greater with UHR than with standard and, compared to standard, the dose reduction potential with UHR was -32.9 ± 0.0 (SD) % for abdominal lesions and -68.7 ± 3.2 (SD) % for bone lesions.

Compared to the standard mode, the UHR mode offers lower noise levels and better detectability of abdominal and bone lesions, paving the way for potential dose reduction with PCCT in clinical applications.