Lung Nodule Volume Quantification and Shape Differentiation with an Ultra-High Resolution Technique on a Photon...

Area of Science:

• Medical Imaging
• Radiology
• Computed Tomography

Background:

• Photon counting-detector (PCD) CT systems offer advanced imaging capabilities.
• Ultra high-resolution (UHR) imaging provides enhanced spatial detail compared to conventional modes.

Purpose of the Study:

• To evaluate the performance of a new UHR mode on a PCD CT system for lung nodule characterization.
• To compare the accuracy of volume estimation and shape differentiation between UHR and conventional (macro) modes.

Main Methods:

• Synthetic lung nodules with varying shapes, sizes, and radio-densities were scanned using UHR and macro modes.
• Images were reconstructed with different kernels, and linear regression analyzed volume accuracy.
• Surface curvature and ROC analysis were used for shape differentiation.

Main Results:

• UHR mode demonstrated more accurate nodule volume estimation, especially for small (3-5mm) and star-shaped nodules, with lower mean absolute percent error (6.5% vs. 11.1-12.9%).
• UHR mode with a sharp reconstruction kernel (S80f) achieved the highest performance (AUC = 0.85) in differentiating star from sphere nodules.
• A strong linear relationship was observed between measured and reference nodule volumes for both modes.

Conclusions:

• The UHR mode on PCD CT scanners significantly improves lung nodule volume measurement accuracy and shape characterization.
• This high-resolution capability offers substantial benefits for quantitative imaging and clinical applications in lung nodule assessment.