Large zoom ratio and adaptive aberration correction microscope using 4DPSF-aware Physical Degradation-guided Network

Area of Science:

• Microscopy
• Optical Engineering
• Computational Imaging

Background:

• Liquid lenses offer real-time optical zooming in microscopy but have limited zoom range and introduce aberrations.
• Dynamic aberrations during zooming degrade image quality, hindering applications in biology, medicine, and materials science.

Purpose of the Study:

• To develop a continuous optical zoom microscope with a large zoom ratio and adaptive aberration correction.
• To address limitations of current liquid lens microscopy, improving image quality across a wide zoom range.

Main Methods:

• An end-to-end joint optimization framework integrating optical design and a neural network guided by physical degradation.
• Incorporation of spatially-variant, multi-wavelength, and continuous-magnification 4D Point Spread Function (4D PSF) as physical priors.
• Development of a 4D PSF-aware Physical Degradation-guided Network (4DPSF-PDNet) for adaptive aberration correction.

Main Results:

• Achieved fast and high-quality continuous zoom imaging from 10.6× to 101.4×.
• Successfully adaptively corrected complex aberrations varying with magnification and spatial location.
• Demonstrated suppression of distortions and artifacts with precise correction of dynamic aberrations.

Conclusions:

• The proposed adaptive continuous microscope offers significant promise for dynamic and cross-scale microscopic observation.
• The integration of hardware (zoom objective, liquid lenses) and advanced algorithms (4DPSF-PDNet) overcomes previous limitations.
• This technology has broad potential applications in biological research, medical diagnostics, and materials science.