Self-calibrating Fourier ptychographic microscopy using automatic differentiation.

Area of Science:

• Microscopy
• Computational Imaging
• Optical Physics

Background:

• Fourier ptychographic microscopy (FPM) enables high-throughput imaging by computationally reconstructing images from low-resolution data.
• Accurate calibration of the optical setup is critical for high space-bandwidth product (SPB) image quality in FPM.
• Current FPM reconstruction methods, including those using machine learning or specific calibration algorithms, often require extensive parameter tuning and additional optical system information, limiting practical application.

Purpose of the Study:

• To introduce a novel, self-calibrating FPM reconstruction approach that eliminates the need for extra optical system information.
• To develop a method that simultaneously recovers a complex 2D image and refines physical parameters of the optical system.
• To reduce the calibration burden associated with FPM while maintaining reconstruction effectiveness.

Main Methods:

• A new FPM reconstruction approach utilizing automatic differentiation.
• Simultaneous recovery of a complex 2D image and refinement of optical system physical parameters.
• Validation through numerical simulations and practical application using ultraviolet Fourier ptychographic microscopy (UV-FPM) on cervical cell slides.

Main Results:

• The proposed method achieves self-calibration without requiring additional system information.
• The approach matches the effectiveness of existing FPM recovery techniques.
• Demonstrated practical utility in reconstructing a full field of view of cervical cell slides using UV-FPM.

Conclusions:

• The novel self-calibrating FPM method significantly simplifies the practical implementation of FPM by removing the need for external calibration data.
• This approach offers a more accessible and efficient solution for high-throughput imaging applications in microscopy.
• The technique holds promise for advancing biological and materials science imaging through improved FPM accessibility and performance.