Source and coded aperture joint optimization for compressive X-ray tomosynthesis.

Area of Science:

• Medical Imaging
• Computational Imaging
• X-ray Imaging

Background:

• Compressive X-ray tomosynthesis reconstructs 3D objects from 2D projections using distributed X-ray sources and coded apertures.
• Image reconstruction quality depends on X-ray source locations, incident angles, and coded apertures influencing structured illumination.
• Reducing radiation dose while maintaining image quality is a key challenge in X-ray imaging.

Purpose of the Study:

• To develop a joint optimization approach for X-ray source and coded apertures (SCO) to enhance compressive X-ray tomosynthesis.
• To improve 3D imaging reconstruction accuracy by minimizing sensing matrix coherence.
• To achieve uniform object sensing and reduce convergence error in image reconstruction.

Main Methods:

• A source and coded aperture joint optimization (SCO) approach based on compressive sensing theory was developed.
• The synergy between source patterns, orientations, and coded apertures was utilized to minimize sensing matrix coherence.
• Gradient-based optimization algorithms and regularization methods were employed for improved reconstruction.

Main Results:

• The SCO approach effectively increases optimization freedom compared to individual source or aperture optimization.
• Significant improvements in 3D imaging reconstruction accuracy were achieved.
• Uniform sensing of the object under inspection was facilitated, reducing convergence error.

Conclusions:

• The proposed SCO method offers a superior strategy for optimizing compressive X-ray tomosynthesis systems.
• This approach leads to enhanced 3D imaging reconstruction accuracy and potentially lower radiation doses.
• Joint optimization of X-ray sources and coded apertures is crucial for advancing tomosynthesis techniques.