Image restoration in computed tomography: estimation of the spatially variant point spread function

Area of Science:

• Medical Imaging
• Radiology
• Image Reconstruction

Background:

• Computed tomography (CT) imaging quality is influenced by the system's point-spread-function (PSF).
• Traditional methods for PSF determination involve experimental scanning of resolution phantoms, which can be resource-intensive.

Purpose of the Study:

• To develop and validate an indirect method for estimating the point-spread-function (PSF) in CT systems.
• To provide an alternative to direct experimental PSF measurement.

Main Methods:

• Modeling the two-dimensional X-ray beam profile as a convolution of focal spot intensity and detector sensitivity.
• Simulating the CT data collection process using known X-ray beam profiles.
• Applying conventional image reconstruction techniques to derive the PSF from simulated data.

Main Results:

• The indirect method successfully estimates CT system PSF by modeling X-ray beam profiles.
• Model validation confirmed accuracy through experimental focal spot intensity measurements.
• Visual comparison showed asymmetry in computed PSFs due to phantom misregistration.

Conclusions:

• An indirect PSF determination method for CT scanners is feasible and offers an alternative to experimental approaches.
• Accurate modeling of X-ray beam profiles is crucial for reliable PSF estimation.
• Potential for improving CT image quality assessment and correction.