Resolution and noise trade-off analysis for volumetric CT

Area of Science:

• Medical Imaging
• Physics

Background:

• Previous studies on spatial resolution vs. quantum noise focused on single-slice CT.
• The impact of reconstruction kernel shapes on this trade-off was not fully explored.

Purpose of the Study:

• Extend theoretical framework for spatial resolution-quantum noise trade-off to volumetric CT.
• Incorporate actual reconstruction kernel shapes into the analysis.
• Develop and apply a 3D spatial resolution measurement technique.

Main Methods:

• Theoretical analysis of noise variance and spatial resolution.
• Experimental measurements using a sphere phantom and a modified FDK algorithm.
• 3D Modulation Transfer Function (MTF) measurement along the spherical surface.

Main Results:

• Theoretical analysis predicted noise variance proportional to the 4th power of spatial resolution.
• Experimental results confirmed a 4.6th (helical) or 5th (axial) power relationship.
• Matched photon flux and consistent reconstruction parameters ensured reliable comparisons.

Conclusions:

• The spatial resolution-quantum noise trade-off in volumetric CT is more complex than previously modeled.
• The findings provide a quantitative basis for optimizing CT imaging parameters.
• This study highlights the importance of considering reconstruction kernel shapes and employing 3D measurement techniques.