An interior point iterative maximum-likelihood reconstruction algorithm incorporating upper and lower bounds with...

Area of Science:

• Medical Imaging
• Computational Science

Background:

• Transmission reconstruction is crucial for accurate imaging.
• Iterative algorithms can struggle with truncated projections, leading to inaccurate attenuation coefficients.

Purpose of the Study:

• To introduce and evaluate the Block-Iterative Interior Point (BITAB) algorithm for transmission reconstruction.
• To assess BITAB's ability to improve attenuation map accuracy using bounded constraints.

Main Methods:

• Developed a block-iterative version of the interior point algorithm for transmission reconstruction.
• Implemented constraints on pixel values (a(j) < x(j)k < b(j)) defining the BITAB method.
• Conducted computer simulations using a 3D cardiac and torso phantom with truncated fan beam projections.

Main Results:

• The BITAB algorithm demonstrated potential for improved accuracy in reconstructed attenuation coefficients.
• Reasonably selected upper and lower bounds effectively restricted overestimation outside fully sampled regions.
• BITAB showed advantages over traditional methods like maximum-likelihood gradient type algorithms for truncated data.

Conclusions:

• The BITAB method offers a promising approach for accurate transmission reconstruction, especially with incomplete projection data.
• Constrained optimization within the BITAB framework is key to mitigating artifacts from truncated fan beam projections.
• Further research into bound selection could optimize BITAB's performance in various imaging scenarios.