A new phase consistency criterion and its application in electron crystallography

Area of Science:

• Image processing
• Crystallography
• Fourier analysis

Background:

• Assessing Fourier components is crucial for image analysis and 3D reconstruction.
• Traditional methods like Signal-to-Noise Ratio (SNR)-based quality index (IQ) have limitations in evaluating image reflections.
• Robust alignment is essential for accurate 3D reconstruction in fields like electron crystallography.

Purpose of the Study:

• To present a novel methodology for assessing Fourier components based on phase consistency.
• To introduce a new phase consistency criterion for spatially translated images.
• To demonstrate the application of this criterion in improving the alignment stage of 3D reconstruction.

Main Methods:

• Defined a new phase consistency criterion leveraging the spatial shift property of the Fourier transform.
• Applied the criterion to sets of spatially translated images.
• Utilized the criterion in the alignment stage of 3D reconstruction, focusing on projection image alignment in electron crystallography.

Main Results:

• The new phase consistency criterion effectively assesses frequency components and enhances alignment robustness.
• Application to electron crystallography image reflections showed the criterion complements traditional SNR-based IQ.
• The criterion successfully identified and allowed discarding of unreliable reflections, improving alignment reliability.

Conclusions:

• The proposed phase consistency definition offers a valuable tool for evaluating Fourier components.
• This method enhances the reliability of image alignment in 3D reconstruction processes.
• It provides a complementary quality index to traditional methods, particularly beneficial in electron crystallography.