Three-dimensional simulations of multidetector point-focusing SPECT imaging

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Computational Science

Background:

• Single-photon emission computed tomography (SPECT) is crucial for functional imaging.
• Accurate simulation of SPECT systems is essential for image reconstruction development.
• Depth-dependent point spread functions (3-DPSF) significantly impact image quality.

Purpose of the Study:

• To develop and apply an efficient algorithm for simulating the 3-D response of SPECT imaging systems.
• To evaluate the impact of a depth-dependent 3-D point spread function (3-DPSF) on image reconstruction.
• To identify and characterize artifacts in reconstructed SPECT images of the human brain.

Main Methods:

• Developed an efficient algorithm for simulating 3-D SPECT system response using depth-dependent 3-D point spread functions (3-DPSF).
• Utilized a sequence of 3-D line spread function (3-DLSF) additions for 3-D convolution.
• Simulated projection data from a multidetector SPECT system with point-focusing collimators.
• Reconstructed simulated data using manufacturer's software.

Main Results:

• The simulation algorithm accurately modeled SPECT imaging with 3-DPSF.
• Reconstruction of simulated data, including realistic brain slices, revealed significant artifacts.
• Both qualitative and quantitative artifacts were observed in the reconstructed human brain images.

Conclusions:

• The developed simulation algorithm is effective for evaluating SPECT imaging systems.
• Depth-dependent 3-DPSF can lead to substantial artifacts in reconstructed SPECT brain images.
• These findings highlight the need for improved image reconstruction algorithms to mitigate artifacts in SPECT neuroimaging.