Deriving the modulation transfer function of CT from extremely noisy edge profiles.

Area of Science:

• Medical Imaging
• Radiology
• Image Quality Assessment

Background:

• The point spread function (PSF) method is standard for X-ray CT modulation transfer function (MTF) but requires high contrast-to-noise ratio (CNR) and zooming reconstruction, limiting clinical applicability.
• The edge spread function (ESF) method avoids zooming reconstruction but is sensitive to image noise, similar to the PSF method.

Purpose of the Study:

• To develop a noise-robust edge spread function (ESF) method for determining the modulation transfer function (MTF) of X-ray CT systems under clinical conditions.
• To enable accurate MTF assessment even with low contrast-to-noise ratios (CNR) and without specialized reconstruction techniques.

Main Methods:

• A novel technique to enhance the robustness of the ESF method against image noise by applying multi-stage filtering to the noisy ESF.
• Correction of blurring introduced in the line spread function (LSF) derived from the smoothed ESF, by deconvolution assuming Gaussian or multi-Gaussian line spread functions.
• Validation of the method using images reconstructed with a soft-tissue imaging kernel.

Main Results:

• The proposed ESF method effectively mitigates noise, allowing for reliable MTF estimation.
• Accurate MTF values were obtained even with a low contrast-to-noise ratio (CNR) of 2.
• The MTF derived from the noise-robust ESF method showed sufficient consistency with theoretical MTF values.

Conclusions:

• The developed ESF-based technique offers a practical and robust alternative for X-ray CT MTF assessment in clinical settings.
• This method overcomes the limitations of the PSF method and improves upon the standard ESF approach by effectively handling image noise.
• Accurate modulation transfer function (MTF) determination is achievable even under challenging low contrast-to-noise ratio (CNR) conditions.