Optimized reconstruction algorithm for helical CT with fractional pitch between 1PI and 3PI

Area of Science:

• Medical Imaging
• Computational Imaging
• Image Reconstruction

Background:

• Exact Katsevich reconstruction algorithms are crucial for medical imaging.
• Utilizing redundant data outside the 1PI window can improve reconstruction accuracy.
• Previous methods have limitations in handling data outside the primary reconstruction window.

Purpose of the Study:

• To develop an approximate approach for incorporating redundant data outside the 1PI window into the exact Katsevich reconstruction framework.
• To enable flexible selection of helical pitch for improved clinical applicability.
• To enhance noise and artifact reduction in tomographic reconstruction.

Main Methods:

• Extending the work of KOhler, Bontus, and Koken (2006).
• Optimizing contribution weights of convolution families within the Katsevich 3PI algorithm.
• Solving a constrained least squares problem to ensure Radon plane weights approximate 1.
• Utilizing redundant data outside the 1PI window.

Main Results:

• Demonstrated good noise reduction properties.
• Showcased effective artifact reduction.
• Validated the algorithm's performance through numerical evaluation.
• The flexible helical pitch selection proved beneficial for clinical scenarios.

Conclusions:

• The proposed approximate approach effectively integrates redundant data into the exact Katsevich framework.
• The algorithm offers significant improvements in noise and artifact reduction.
• The flexible helical pitch selection enhances clinical utility in medical imaging applications.