Development and Evaluation of Image Reconstruction Algorithms for a Novel Desktop SPECT System

Area of Science:

• Nuclear Medicine
• Medical Imaging
• Image Reconstruction

Background:

• Iterative reconstruction algorithms are crucial for modern nuclear medicine imaging systems.
• Selecting and evaluating appropriate algorithms is essential for optimizing image quality.
• Novel imaging systems require tailored reconstruction approaches for performance assessment.

Purpose of the Study:

• To develop and compare three image reconstruction algorithms for the PERSPECT desktop open-gantry SPECT system.
• To assess the performance of qSMART, MLEM, and OSEM in terms of reconstructed image quality.
• To determine the most effective algorithm for the novel SPECT system.

Main Methods:

• Implemented and compared quasi-simultaneous multiplicative algebraic reconstruction technique (qSMART), maximum-likelihood expectation-maximization (MLEM), and ordered-subsets EM (OSEM).
• Utilized analytic and Monte Carlo simulations with various phantoms, including a micro-Derenzo phantom.
• Quantitatively evaluated image quality using a comprehensive set of figures-of-merit across 1-40 iterations.

Main Results:

• OSEM consistently produced higher or equivalent image quality compared to qSMART.
• MLEM achieved comparable quality to OSEM but required more iterations.
• Convergence speed ranked as qSMART, OSEM, MLEM; OSEM yielded optimal bias-noise trade-offs.

Conclusions:

• While qSMART is viable for early iterations (up to 5), OSEM delivers the best achievable image quality.
• OSEM is the recommended algorithm for the PERSPECT system based on image quality and bias-noise performance.
• The study highlights the importance of algorithm selection for SPECT imaging performance.