Estimation of the Optimal Iteration Number for Minimal Image Discrepancy

Area of Science:

• Medical Imaging
• Computational Science

Background:

• Iterative image reconstruction algorithms are susceptible to noise, necessitating early termination before full convergence.
• Determining the optimal stopping point to minimize reconstruction error remains an open problem in the field.

Purpose of the Study:

• To address the open problem of identifying the optimal stopping point in iterative image reconstruction.
• To establish novel approximate relationships for estimating this optimal stopping point.

Main Methods:

• Establishing an approximate relationship between the iterative Landweber algorithm and an iteration-number-emulated filtered backprojection (FBP) algorithm.
• Establishing a second approximate relationship between optimal iteration-number-emulated FBP reconstruction and optimal projection-domain filtered data.

Main Results:

• Two approximate relationships were successfully established, linking iterative methods with filtered backprojection and projection-domain data.
• These relationships provide a pathway to estimate the optimal iteration count for image reconstruction.

Conclusions:

• The established relationships offer a promising approach to estimating the optimal stopping point in iterative image reconstruction.
• This work contributes to improving the accuracy and reliability of medical image reconstruction techniques by mitigating noise effects.