Metasurface-based Fourier ptychographic microscopy

Area of Science:

• Optics and Photonics
• Computational Imaging
• Biomedical Engineering

Background:

• Meta-optics enable miniaturized, high-performance imaging systems.
• Fourier ptychography (FP) offers high-resolution quantitative phase imaging but traditionally requires bulky optics and mechanical scanning.
• Existing FP methods face challenges in portability and speed.

Purpose of the Study:

• To develop a compact and fast Fourier ptychography (FP) microscopy platform.
• To overcome the limitations of traditional FP systems using meta-optics and artificial intelligence.
• To demonstrate high-resolution quantitative phase imaging with improved portability and speed.

Main Methods:

• Integration of a 4-f metalens system for miniaturization.
• Utilized a programmable thin-film transistor (TFT) panel for angle-diverse illumination without mechanical movement.
• Employed a residual convolutional neural network (RCNN) for single-shot, high-resolution phase reconstruction from low-resolution inputs.

Main Results:

• Achieved nearly twofold resolution improvement (7.81 µm to 3.91 µm).
• Demonstrated accurate quantitative phase recovery with <10% error on phase standards.
• Showcased reliable dry-mass estimation of H1975 cells with <12% average deviation.

Conclusions:

• The meta-FP platform offers a fast, compact, and high-performance solution for quantitative phase imaging.
• Integration of metasurface optics and AI accelerates imaging and enhances resolution.
• Potential applications include live-cell imaging, microfluidic monitoring, and point-of-care diagnostics.