Physics-based iterative reconstruction for dual-source and flying focal spot computed tomography.

Area of Science:

• Medical Imaging
• Computed Tomography (CT)
• Image Reconstruction

Background:

• Filtered-Back Projection (FBP) and statistical iterative reconstruction are established for single-source CT.
• Accurate statistical iterative reconstruction for dual-source CT with flying focal spot (DS-FFS CT) is lacking.
• Existing methods do not fully model DS-FFS CT scanner geometry and acquisition physics.

Purpose of the Study:

• To present a novel physics-based iterative reconstruction method specifically for DS-FFS CT.
• To assess the image quality achieved by the proposed reconstruction method.

Main Methods:

• Developed a reconstruction algorithm using precise physics models for DS-FFS CT's cone-beam geometry and helical trajectory.
• Incorporated noise physics and prior image models to represent noise and texture.
• Employed advanced Consensus Equilibrium (CE) numerical methods for image reconstruction via a joint optimization problem.

Main Results:

• The novel reconstruction demonstrated higher spatial resolution than the clinical standard Siemens ADMIRE (hybrid iterative reconstruction).
• Task-Based Modulation Transfer Function (MTF_task) was consistently higher with the new method.
• Reduced image undersampling artifacts were observed compared to the clinical standard.

Conclusions:

• The physics-based reconstruction accurately models DS-FFS CT geometry without data rebinning or interpolation.
• Achieved superior spatial resolution and significantly fewer, lower-magnitude image artifacts.
• Represents an advancement over current clinical standard hybrid iterative reconstruction for DS-FFS CT.