ADMM approach for efficient iterative tomographic deconvolution reconstruction of 3D quantitative phase images.

Area of Science:

• Optics and Photonics
• Biomedical Imaging
• Computational Imaging

Background:

• Tomographic deconvolution phase microscopy (TDPM) enables 3D quantitative imaging of phase objects.
• Biological cells and optical fibers are key phase objects requiring precise imaging.
• Existing TDPM methods face limitations in acquisition and processing speed and accuracy.

Purpose of the Study:

• To enhance Tomographic Deconvolution Phase Microscopy (TDPM) using the Alternating Direction Method of Multipliers (ADMM).
• To reduce image acquisition and processing times for TDPM.
• To improve the accuracy of 3D quantitative imaging of phase objects.

Main Methods:

• Application of the Alternating Direction Method of Multipliers (ADMM) to TDPM.
• Optimization of image fidelity through Gaussian noise minimization.
• Implementation of total variation regularization with nonnegativity and known zero constraints.
• Reconstruction of shift-variant phase objects in three spatial dimensions.

Main Results:

• ADMM-TDPM achieved a 5x speedup in image acquisition time.
• ADMM-TDPM demonstrated over 10x speedup in image processing time.
• The enhanced method resulted in higher accuracy compared to conventional TDPM.

Conclusions:

• ADMM-TDPM offers a significant advancement in 3D quantitative phase imaging.
• The method provides substantial improvements in speed and accuracy for TDPM.
• ADMM-TDPM is a powerful tool for imaging dynamic and complex phase objects.