Density resolution in quantitative computed tomography of foam and lung

Area of Science:

• Medical Imaging
• Materials Science

Background:

• Quantitative computed tomography (CT) assesses material density using CT number histograms.
• Density resolution, crucial for discriminating materials, is typically limited by quantum noise.
• Cellular solids, like lung tissue and foams, introduce partial volume effects that further degrade density resolution.

Purpose of the Study:

• To evaluate density resolution in quantitative CT of foam models and human lung tissue.
• To investigate the impact of sample volume and spatial resolution on density resolution.

Main Methods:

• Utilized polyethene (PE) foams as models for lung tissue.
• Examined five patients using varying section thicknesses and reconstruction filters.
• Determined density resolution using CT number histogram analysis (Full Width at Half Maximum for foams).
• Applied a combined convolution least-squares fit for patient data analysis.

Main Results:

• For foams, density resolution was primarily limited by sample size, not quantum noise, particularly with cell sizes of 0.8-1.5 mm.
• Density resolution in both foams and lung tissue showed strong dependence on sample volume.
• Poor density resolution was observed with thin sections and sharp filters.
• Histogram-shape parameters were sensitive to the chosen spatial resolution.

Conclusions:

• Thin section densitometry (1-mm) with standard or high-resolution filters is not recommended for accurate density analysis, except for average density determination.
• Thicker sections are advised for improved density resolution in CT imaging of lung and similar materials.
• Matching in-plane spatial resolution to section thickness is recommended for optimal results.