A cone-beam filtered backprojection reconstruction algorithm for cardiac single photon emission computed tomography.

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Computational Imaging

Background:

• Cardiac single photon emission computed tomography (SPECT) is crucial for diagnosing heart conditions.
• Cone-beam geometry presents unique reconstruction challenges compared to traditional fan-beam SPECT.
• Photon attenuation significantly degrades image quality in SPECT reconstructions.

Purpose of the Study:

• To develop a filtered backprojection algorithm for cardiac SPECT using cone-beam geometry.
• To address and minimize attenuation artifacts in cardiac SPECT imaging.
• To evaluate the impact of attenuation versus cone-beam sampling artifacts on image reconstruction.

Main Methods:

• Developed a filtered backprojection algorithm for cone-beam SPECT.
• Implemented a 'short scan' acquisition strategy on a noncircular planar orbit.
• Simulated cardiac SPECT data to assess reconstruction performance.

Main Results:

• The algorithm reconstructs data from short scan acquisitions, minimizing posterior projection attenuation.
• Computer simulations indicate attenuation artifacts have a greater impact on quantitation and image quality than cone-beam sampling artifacts.
• Reconstruction errors in the heart were minimal (≤2%) with nonattenuated projection data.

Conclusions:

• The developed algorithm effectively reconstructs cardiac SPECT data in cone-beam geometry.
• Minimizing photon attenuation is critical for accurate cardiac SPECT imaging.
• The approach shows promise for improving cardiac SPECT image quality and diagnostic accuracy.