Half-scan fan-beam computed tomography with improved noise and resolution properties

Area of Science:

• Medical Imaging
• Image Reconstruction
• Computed Tomography

Background:

• Half-scan strategy in fan-beam computed tomography (CT) reduces scan time and radiation dose.
• Standard half-scan filtered backprojection (FFBP) algorithms can suffer from nonuniform resolution and noise due to aliasing and data noise.
• Uniform image properties are crucial for accurate estimation, detection, and classification tasks.

Purpose of the Study:

• To develop and evaluate a novel algorithm for image reconstruction in half-scan CT.
• To achieve uniform spatial resolution and noise properties in reconstructed images.
• To provide analytic expressions for image variances to assess noise properties.

Main Methods:

• Development of a new image reconstruction algorithm for half-scan CT.
• Derivation of analytic expressions for image variances to evaluate noise properties.
• Quantitative assessment of resolution and noise using numerical studies.

Main Results:

• The proposed algorithm produces images with more uniform spatial resolution compared to the half-scan FFBP algorithm.
• The new algorithm results in lower and more uniform noise levels in reconstructed images.
• Empirical results validate the derived analytic expressions for image variances.

Conclusions:

• The proposed algorithm effectively addresses the limitations of traditional FFBP for half-scan CT.
• This method is particularly beneficial for CT configurations with short focal lengths and large fields of measurement, such as micro-CT and radiation therapy CT.
• The derived analytic expressions aid in assessing image quality for detection and classification tasks.