Low-dosage Maximum-A-Posteriori Focusing and Stigmation

Area of Science:

• Electron microscopy
• Image processing
• Materials science

Background:

• High-resolution imaging is often limited by radiation damage.
• Low-dose focusing and stigmation are critical for preserving sample integrity.
• Automated, long-term imaging techniques like serial block-face electron microscopy require optimal resolution throughout acquisition.

Purpose of the Study:

• To develop an algorithm for optimal signal utilization in low-dose electron microscopy.
• To improve focusing and stigmation accuracy under radiation dose constraints.
• To enhance image quality in demanding imaging applications.

Main Methods:

• Developed the Maximum-A-Posteriori Focusing and Stigmation (MAPFoSt) algorithm.
• MAPFoSt estimates focus and astigmatism using two test images at different focus settings.
• Algorithm designed for optimal use of available signal, even at low electron doses.

Main Results:

• MAPFoSt outperforms commercial scanning electron microscope focusing algorithms at tenfold reduced dose.
• Accurate estimation of focus and astigmatism coefficients (<7% of depth of focus) achieved with low electron dose (2,500 electrons/pixel) and low signal-to-noise ratio.
• Demonstrated effective aberration correction with minimal detected signal (19 secondary electrons/pixel).

Conclusions:

• MAPFoSt enables high-quality imaging with significantly reduced radiation dose.
• The algorithm is particularly beneficial for applications requiring sustained optimal resolution, such as serial block-face electron microscopy.
• Potential for generalization to higher-order aberrations and diverse 2D/3D imaging modalities.