An efficient reconstruction method for nonuniform attenuation compensation in nonparallel beam geometries based on...

Area of Science:

• Medical Imaging
• Image Reconstruction
• Computational Science

Background:

• Accurate image reconstruction is crucial for quantitative single photon emission computed tomography (SPECT).
• Nonparallel beam (NPB) geometries, often used in cardiac imaging, pose reconstruction challenges.
• Existing methods may struggle with varying focal lengths and noise.

Purpose of the Study:

• To develop and validate an accurate reconstruction method for the attenuated Radon transform in NPB geometries.
• To enhance spatial resolution and computational efficiency in image reconstruction.
• To facilitate quantitative SPECT imaging, particularly for cardiac applications.

Main Methods:

• The method involves three key steps: 1D phase-shift rebinning, nonuniform Hilbert transform, and Novikov's explicit inversion formula.
• The approach is adaptive to various fan-beam geometry settings, from long to short focal lengths.
• Comparison with conventional bilinear rebinning was performed using modulation transfer function (MTF).

Main Results:

• The proposed method demonstrated superior spatial resolution compared to bilinear rebinning.
• Numerical experiments confirmed the method's computational efficiency and stability under Poisson noise.
• Reconstruction quality in complex NPB geometries approached that of parallel-beam geometry.

Conclusions:

• The developed reconstruction method accurately inverts the attenuated Radon transform in NPB geometries.
• This technique offers improved spatial resolution and robustness to noise and geometric variations.
• The method is suitable for quantitative SPECT cardiac imaging, addressing the need for NPB collimation and high computational efficiency.