GAN-based quantitative oblique back-illumination microscopy enables computationally efficient epi-mode refractive...

Area of Science:

• Biomedical Optics
• Imaging Science
• Computational Imaging

Background:

• Quantitative oblique back-illumination microscopy (qOBM) offers 3D refractive index (RI) tomography for scattering samples.
• Current qOBM reconstruction methods are computationally intensive, limiting real-time applications.

Purpose of the Study:

• To develop a faster and high-fidelity 3D RI tomography method using qOBM.
• To leverage generative adversarial networks (GANs) for improved qOBM reconstruction.

Main Methods:

• A novel GAN-assisted approach was developed for qOBM 3D RI tomography.
• The method utilizes differential phase contrast images from three adjacent z-planes.
• Reconstruction algorithms were optimized for speed and accuracy.

Main Results:

• Achieved a ~9-fold reduction in volumetric reconstruction time.
• Demonstrated high-fidelity 3D RI tomograms with high signal-to-noise ratio (SNR).
• The technique showed robustness across different tissue types and reduced motion artifacts.

Conclusions:

• The GAN-assisted qOBM approach significantly accelerates 3D RI tomography.
• This method offers high quantitative fidelity and generalizes well to various tissues.
• Enables potential for real-time, in-vivo pre-clinical and clinical imaging.