Combining ptychographical algorithms with the Hybrid Input-Output (HIO) algorithm.

Area of Science:

• Computational imaging
• Phase retrieval algorithms
• X-ray microscopy

Background:

• Ptychographical Iterative Engine (PIE) is a widely used phase retrieval algorithm.
• The Hybrid Input-Output (HIO) algorithm is often employed to refine reconstructions.
• Integrating feedback mechanisms in PIE can improve results, but requires careful implementation.

Purpose of the Study:

• To introduce a novel method for combining the Ptychographical Iterative Engine (PIE) with the Hybrid Input-Output (HIO) algorithm.
• To enhance the performance of ptychographical phase retrieval by optimizing the HIO feedback mechanism.
• To demonstrate the general applicability of the proposed method to other ptychographical algorithms.

Main Methods:

• The core innovation involves maintaining strict separation between the HIO feedback function and the reconstructed object.
• A distinct feedback function is introduced for each probe position.
• The method was tested by combining HIO with floating PIE (fPIE) and extended PIE (ePIE).

Main Results:

• Simulations show that the combined algorithm achieves significantly better performance across various scenarios.
• The separation of the HIO feedback function leads to more robust and accurate phase reconstructions.
• The enhanced method demonstrates improved convergence and reduced artifacts in simulated ptychographic data.

Conclusions:

• The proposed method of separating the HIO feedback function offers a significant advancement in ptychographical phase retrieval.
• This technique can be readily adapted to improve other ptychographical algorithms that utilize feedback functions.
• The findings suggest a new standard for integrating feedback mechanisms in iterative phase retrieval processes.