A systematic assessment and optimization of photon-counting CT for lung density quantifications

Area of Science:

• Radiological imaging physics
• Medical imaging technology
• Pulmonary diagnostics

Background:

• Photon-counting computed tomography (PCCT) is a new clinical imaging modality.
• Optimal imaging protocols and benefits of PCCT for lung density quantification are not yet fully understood.
• Comparison with energy-integrating computed tomography (EICT) is needed.

Purpose of the Study:

• To systematically evaluate the performance of a clinical PCCT scanner for lung density quantification.
• To compare the accuracy of PCCT lung density measurements against EICT scanners.

Main Methods:

• Retrospective analysis of clinical PCCT scans and virtual CT datasets of anthropomorphic subjects.
• Systematic evaluation of reconstruction parameters (kernel, pixel size, slice thickness) on PCCT image quality.
• Comparison of mean absolute error (MAE) for lung and emphysema density against ground truth using PCCT and EICT simulators.

Main Results:

• Sharper kernels and smaller voxel sizes in PCCT increased low-attenuation areas.
• Optimized PCCT settings (higher dose, thinner slices, smaller pixels, iterative reconstruction, medium sharpness kernels) reduced lung MAE.
• PCCT demonstrated lower lung and emphysema MAE compared to both EICT scanners at optimal settings.

Conclusions:

• PCCT shows superior performance in lung density quantification compared to EICT.
• Imaging parameter influences on PCCT are consistent with EICT.
• PCCT offers improved accuracy for objective assessment of respiratory conditions.