Reducing negativity artifacts in emission tomography: post-processing filtered backprojection solutions.

Area of Science:

• Medical Imaging
• Image Reconstruction
• Tomography

Background:

• Negative artifacts in filtered backprojection (FBP) emission tomography reconstructions are problematic, especially in low-count studies.
• Maximum likelihood (ML) methods avoid negativity but are computationally intensive.

Purpose of the Study:

• To develop a computationally efficient post-processing technique to eliminate negative artifacts in FBP reconstructions.
• To evaluate the performance of the proposed method against standard FBP and ML algorithms.

Main Methods:

• A novel iterative post-processing technique was developed to adjust negative values by local neighborhood cancellation.
• Numerical simulations in 1D and 2D using standard phantoms (Hoffman, Shepp-Vardi, Jaszczak) were performed.
• Comparisons included smoothed FBP, post-processed FBP, and expectation-maximization ML, with RMS error and region-of-interest measures.

Main Results:

• The post-processing method demonstrated significant reduction of negative artifacts.
• Computational time was comparable to two FBP applications, much faster than ML.
• Optimized smoothing was critical for minimizing root mean square error across all methods.

Conclusions:

• The proposed post-processing technique offers a practical solution for negative artifacts in FBP emission tomography.
• It provides a computationally feasible alternative to ML methods, particularly beneficial for low-count imaging scenarios.