Iterative k-t principal component analysis with nonrigid motion correction for dynamic three-dimensional cardiac...

Area of Science:

• Medical Imaging
• Cardiovascular Imaging
• Image Reconstruction

Background:

• Dynamic contrast-enhanced (DCE) perfusion imaging provides crucial information on myocardial blood flow.
• Respiratory motion artifacts degrade image quality and complicate quantitative analysis in 3D DCE perfusion imaging.
• Accurate assessment of myocardial perfusion is vital for diagnosing and managing cardiovascular diseases.

Purpose of the Study:

• To propose an iterative k-t principal component analysis (PCA) algorithm for motion correction in 3D DCE perfusion imaging.
• To address challenges posed by respiratory motion during free-breathing or interrupted breath-holds.
• To enhance the accuracy and reliability of myocardial perfusion quantification.

Main Methods:

• An iterative k-t PCA algorithm was developed with regularization using motion-corrected training data.
• Shape-constrained nonrigid image registration was employed to extract motion information.
• The method was validated using computer simulations and in vivo data with 10-fold k-t undersampling, comparing results to breath-held references.

Main Results:

• The proposed k-t PCA algorithm with motion correction effectively resolved aliasing artifacts.
• Signal intensity curves from the myocardium closely matched those from breath-held reference data.
• Myocardial upslope measurements demonstrated improved spatial homogeneity with the motion-corrected k-t PCA approach.

Conclusions:

• Iterative k-t PCA combined with nonrigid motion correction successfully corrects respiratory motion artifacts in 3D first-pass myocardial perfusion imaging.
• This technique offers a promising solution for improving diagnostic accuracy in free-breathing DCE cardiac MRI.
• The method enhances the robustness of quantitative perfusion analysis in the presence of motion.