Level-set reconstruction algorithm for ultrafast limited-angle X-ray computed tomography of two-phase flows.

Area of Science:

• Medical Imaging
• Computational Physics
• Fluid Dynamics

Background:

• Tomographic image reconstruction recovers object details from projection data.
• Limited angular views (<180°) in X-ray computed tomography cause reconstruction artifacts with standard algorithms.
• Specialized algorithms incorporating prior information are needed for limited-angle reconstruction.

Purpose of the Study:

• To develop a novel reconstruction algorithm for ultrafast limited-angle X-ray computed tomography of two-phase flows.
• To address the challenge of reconstructing binary distributions in two-phase systems.
• To improve accuracy and reduce artifacts in limited-angle tomographic reconstructions.

Main Methods:

• A new level-set reconstruction (LSR) algorithm was developed.
• The LSR algorithm incorporates projection data matching and curvature-dependent smoothing.
• Algorithm performance was validated using simulated and measured projection data.

Main Results:

• The LSR algorithm effectively reconstructs binary distributions of two phases.
• Validation demonstrated the algorithm's accuracy compared to algebraic reconstruction techniques.
• The LSR algorithm showed advantages over existing methods in dynamic two-phase flow applications.

Conclusions:

• The proposed level-set reconstruction (LSR) algorithm is applicable to ultrafast limited-angle X-ray computed tomography.
• LSR offers significant advantages over traditional methods for reconstructing two-phase flows.
• This method aids in analyzing dynamic two-phase flow systems with improved accuracy.