Factors affecting the correlation coefficient template matching algorithm with application to real-time 2-D coronary...

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Cardiovascular Imaging

Background:

• Accurate motion tracking is crucial for analyzing coronary artery dynamics.
• Template matching algorithms, like correlation coefficient (CC) template matching, are used for motion tracking.
• Real-time coronary artery magnetic resonance imaging (CMR) presents unique challenges for motion tracking accuracy.

Purpose of the Study:

• To characterize factors influencing the accuracy of CC template matching for coronary artery motion tracking.
• To analyze the algorithm's performance under random and systematic error conditions.
• To develop and validate techniques for enhancing motion tracking accuracy in CMR.

Main Methods:

• Analysis of CC template matching algorithm performance with varying error types.
• Identification and quantification of factors affecting the CC difference-to-noise ratio (CCDNR).
• Development of novel techniques to mitigate systematic errors and optimize CCDNR.
• Validation using phantom studies and real coronary artery MR images.

Main Results:

• A high CCDNR is a necessary and sufficient condition for accurate motion tracking under random error.
• CCDNR is influenced by image/template size, structure, and noise magnitude.
• Large CCDNR alone is insufficient for accuracy with systematic error; new techniques are required.
• Developed methods successfully improved motion tracking accuracy in phantoms and clinical images.

Conclusions:

• Optimizing CCDNR through manipulation of image/template factors enhances accuracy in the presence of random error.
• Specific techniques are effective in minimizing systematic error effects while maintaining sufficient CCDNR.
• The proposed methods significantly improve the reliability of motion tracking in real-time coronary artery MRI.