Author: Azza Mohamed Ahmed, Nadine Francis, Osama Khan, Nabil Maalej, Aamir Raja, Briya Tariq 👨🔬
Affiliation: Khalifa University 🌍
Purpose: The study aims to evaluate the task-specific diagnostic performance of spectral photon-counting CT (SPCCT) using application-specific phantoms.
Methods:
Parameters such as linearity response, modulation transfer function (MTF), signal-to-noise ratio (SNR), and noise power spectrum (NPS) were evaluated. Mars Microlab scanner (5x120) was used with two multi-energy threshold acquisition protocols: one for crystal arthropathies (120 kVp and 60 µA; ETH =12, 15, 18, and 21 keV) and the other for bone characterization in the presence of metal artefacts (120 kVp and 80 µA; ETH = 40, 50, 60, and 79 keV). Acquisitions with both protocols were conducted across different numbers of circular projections (373, 981, and 1440) and reconstructed voxel sizes (70, 80, 90, and 100 µm). Spectral linearity response was measured with varying concentrations of hydroxyapatite (HA) using a QRM D100 spectral CT phantom. A dedicated Mars wire phantom and QRM D100 gout crystal phantom were used for spatial resolution, SNR and NPS assessments, respectively. Contrast-detail curves were also generated using gout crystal phantom to evaluate the system’s ability to resolve low-contrast structures using varying concentrations and sizes of HA crystals.
Results: Results show a strong correlation (R² >0.98) for spectral linearity assessments across all energy bins. SNR was reduced by ~30% with reducing voxel sizes (100 to 70 µm) and increased by ~28% with increasing projections (373 to 981); with a significant reduction in NPS (~40%) at 80 µm voxel size. MTF was improved > 23% with reducing voxel size (100 to 70µm). 981 projections with 80 µm voxel size were observed to provide optimal NPS and MTF performance between noise characteristics and spatial resolution.
Conclusion:
In conclusion, task-specific image quality evaluations to optimize spectral acquisition parameters show that SPCCT has the potential for enhanced diagnostic accuracy in crystal arthropathies and bone assessment.