Pixel-wise quantification of myocardial perfusion using spatial Tikhonov regularization

Area of Science:

• Cardiovascular Imaging and Diagnostics
• Medical Physics
• Image Analysis

Background:

• Accurate quantification of myocardial perfusion is crucial for cardiovascular disease risk assessment.
• Current pixel-wise analysis methods in cardiovascular magnetic resonance imaging (CMR) suffer from reduced spatial resolution or unstable fits due to low signal-to-noise ratio.
• Observer dependency and reproducibility remain challenges in current perfusion quantification.

Purpose of the Study:

• To develop a novel pixel-wise analysis method for myocardial perfusion quantification using CMR.
• To improve the accuracy and reproducibility of cardiovascular disease risk assessment.
• To address limitations of existing methods, particularly in low signal-to-noise ratio scenarios.

Main Methods:

• Implementation of a new pixel-wise analysis based on spatial Tikhonov regularization.
• Exploitation of spatial smoothness in cardiac perfusion data.
• Automatic determination of the regularization parameter using an L-curve criterion.
• Validation using a numerical phantom and patient data.

Main Results:

• The proposed spatial Tikhonov regularization significantly reduces root-mean square error in perfusion estimates compared to non-regularized fits.
• The method demonstrates accurate quantification even with low signal-to-noise ratio images.
• Successful recovery of myocardial perfusion and differentiation between healthy and ischemic regions in patient data.

Conclusions:

• Spatial Tikhonov regularization offers an observer-independent and reproducible approach for myocardial perfusion quantification via CMR.
• This method enhances the diagnostic capability of CMR for cardiovascular disease.
• The technique provides robust performance across varying image quality, improving clinical applicability.