Wavefront estimation through structured detection in laser scanning microscopy

Area of Science:

• Optical imaging
• Microscopy
• Biophysics

Background:

• Laser scanning microscopy (LSM) is crucial for life sciences but limited by optical aberrations.
• Adaptive optics (AO) can correct aberrations but often requires invasive hardware or long sample exposure.
• Current aberration estimation methods are sample-invasive and less user-friendly.

Purpose of the Study:

• To propose a simple, efficient adaptive optics (AO) strategy for confocal laser scanning microscopy (CLSM).
• To demonstrate that image-scanning microscopy datasets inherently contain aberration information.
• To enable faster, less invasive, and more accessible high-resolution imaging.

Main Methods:

• Utilized image-scanning microscopy with a detector array and an adaptive optics (AO) element for beam shaping.
• Developed a custom convolutional neural network (CNN) to decode aberration information.
• Estimated aberrations up to the 11th Zernike coefficient from single-acquisition datasets.

Main Results:

• Demonstrated that detector array datasets inherently encode aberration information.
• Successfully decoded aberrations using a CNN from single CLSM acquisitions.
• Established a new paradigm for aberration sensing in LSM.

Conclusions:

• The proposed method enables faster, less invasive aberration estimation in LSM.
• This data-driven approach opens avenues for advanced aberration decoding strategies.
• The strategy synergizes with conventional AO methods, enhancing accessibility to high-resolution imaging.