Related Experiment Video
Updated: Jul 9, 2025
14:09
Quantitative Optical Microscopy: Measurement of Cellular Biophysical Features with a Standard Optical Microscope
Published on: April 7, 2014
15.6K
Polarization differential interference contrast microscopy with physics-inspired plug-and-play denoiser for
Mariia Aleksandrovych1, Mark Strassberg1, Jonathan Melamed2
1Dept. of Physics and Astronomy, Hunter College and the Graduate Center, The City University of New York, 695 Park Ave, New York, NY 10065, USA.
Biomedical Optics Express
|November 29, 2023
Summary
We developed a new method for high-performance quantitative phase imaging using a deep learning denoiser with polarization differential interference contrast microscopy. This technique significantly improves phase retrieval accuracy and image quality for biological samples.
Area of Science:
- Optics and Photonics
- Biomedical Imaging
- Computational Imaging
Background:
- Quantitative phase imaging (QPI) provides label-free contrast for transparent specimens.
- Polarization differential interference contrast (PDIC) microscopy is a QPI technique sensitive to optical path differences.
- Traditional PDIC methods often struggle with noise and limited resolution, hindering quantitative analysis.
Purpose of the Study:
- To develop a single-shot, high-performance quantitative phase imaging method for PDIC microscopy.
- To enhance phase retrieval accuracy and image quality by integrating a physics-inspired deep learning denoiser.
- To validate the performance of the proposed method on simulated data, phantoms, and biological tissue samples.
Main Methods:
- A physics-inspired plug-and-play denoiser, a custom dense residual U-Net (DRUNet) with Tanh activation, was developed for phase retrieval.
- The quantitative phase was recovered using the alternating direction method of multipliers (ADMM), balancing total variance regularization and the DRUNet denoiser.
- An adaptive strategy was incorporated to accelerate convergence and account for measurement noise.
Main Results:
- The deep denoiser-enhanced PDIC microscopy achieved significantly higher quality and accuracy in phase retrieval compared to conventional methods.
- Validation on simulated data and phantom experiments confirmed the method's robustness and effectiveness.
- High-performance phase imaging of histological tissue sections demonstrated the practical applicability of the technique.
Conclusions:
- The integration of a deep learning denoiser with PDIC microscopy offers a powerful approach for high-performance quantitative phase imaging.
- This method overcomes limitations of traditional techniques, providing superior phase retrieval accuracy and image quality.
- The developed technique holds promise for label-free, quantitative analysis of biological specimens in various research and clinical applications.
