Deblurring Computed Tomography Medical Images Using a Novel Amended Landweber Algorithm

Area of Science:

• Medical Imaging
• Image Processing
• Computational Science

Background:

• Computed tomography (CT) images often suffer from blurring caused by hardware or software errors, obscuring critical medical details.
• Image blurring is a common issue that can significantly degrade image quality and hinder accurate diagnosis.
• Existing deblurring methods for CT images face challenges such as long implementation times, noise amplification, and boundary artifacts.

Purpose of the Study:

• To present an amended iterative Landweber algorithm for CT image deblurring.
• To address the limitations of existing methods, specifically high implementation time, noise amplification, and boundary artifacts.
• To achieve artifact-free boundaries and reduced noise amplification in CT images more efficiently.

Main Methods:

• An amended iterative Landweber algorithm was developed and applied to both synthetic and real blurred CT images.
• The proposed method was validated using synthetic and real-world CT image datasets.
• Image quality assessment was performed using pixel domain metrics including feature similarity index (FSI), structural similarity (SSIM), and visual information fidelity (VIF).

Main Results:

• The amended Landweber algorithm demonstrated efficient artifact reduction in CT images.
• The method achieved artifact-free boundaries and exhibited less noise amplification compared to existing techniques.
• Experimental results confirmed the faster application time of the proposed algorithm.

Conclusions:

• The proposed amended Landweber algorithm is effective in deblurring CT images.
• The algorithm offers significant improvements in terms of speed, noise control, and boundary artifact reduction.
• This enhanced method shows potential as a valuable tool for medical image processing applications.