A g-factor metric for k-t SENSE and k-t PCA based parallel imaging

Area of Science:

• Magnetic Resonance Imaging
• Image Reconstruction
• Signal Processing

Background:

• Accelerated MRI techniques like k-t SENSE and k-t PCA reduce scan times but can impact image quality.
• Assessing the trade-offs between acceleration, noise, and temporal fidelity is crucial for clinical applications.
• Quantitative metrics are needed to compare the performance of different spatiotemporal reconstruction methods.

Purpose of the Study:

• To propose and validate a g-factor formalism for evaluating k-t SENSE, k-t PCA, and related methods.
• To assess signal-to-noise ratio (SNR) and temporal fidelity in accelerated MRI.
• To provide a framework for quantitative comparison of k-t reconstruction techniques.

Main Methods:

• Derived an analytical g-factor (gxf) formulation in the spatiotemporal frequency domain.
• Validated the gxf-factor using pseudoreplica analysis on cardiac cine MRI data.
• Analyzed the performance of k-t methods under various undersampling factors and parameter settings.

Main Results:

• Analytical gxf-factor maps showed good agreement with pseudoreplica analysis for k-t SENSE and k-t PCA at 3x, 5x, and 7x acceleration.
• k-t PCA demonstrated better temporal fidelity preservation at higher undersampling factors compared to k-t SENSE.
• Differences in average gxf values were observed between k-t SENSE and k-t PCA in static versus dynamic regions.

Conclusions:

• The proposed gxf-factor and temporal transfer formalism effectively assess noise and temporal fidelity for k-t methods.
• This framework allows for quantitative comparison of k-t SENSE, k-t PCA, and other k-t techniques.
• Enables objective evaluation relative to conventional frame-by-frame parallel imaging reconstruction.