Iterative micro-tomography of biopsy samples from truncated projections with quantitative gray values.

Area of Science:

• Medical Imaging
• Computational Science
• Biotechnology

Background:

• Three-dimensional (3D) reconstruction from limited two-dimensional (2D) projections is an ill-posed problem without additional information.
• Reconstruction artifacts, such as peripheral glow, hinder accurate analysis in local tomography, especially for biological samples.
• Existing methods often require complete projection data or prior knowledge of the sample's structure.

Purpose of the Study:

• To develop an iterative algorithm for artifact suppression in region of interest (ROI) local tomography.
• To improve the quantitative accuracy and image quality of 3D reconstructions from truncated projection data.
• To enable enhanced analysis of micro-tomography data from biological tissue samples.

Main Methods:

• An iterative algorithm based on back-projection and re-projection was devised.
• The algorithm assumes general homogeneity and an approximately cylindrical sample shape.
• Reconstruction refinement involves minimizing the mismatch between an empty ROI and the projected reconstruction from sinogram differences.

Main Results:

• The algorithm successfully suppresses reconstruction artifacts, including peripheral glow.
• Quantitative gray values are accurately reconstructed.
• Demonstrated improvements in image sharpness and reduced artifacts in numerical simulations and experimental data.

Conclusions:

• The iterative algorithm effectively addresses limitations in 3D reconstruction from truncated projections.
• This method is particularly suitable for micro/nano-computed tomography of soft biological tissues, such as biopsy and autopsy samples.
• The technique enhances the diagnostic potential of computed tomography for biological and medical applications.