Cardiac magnetic resonance imaging using radial k-space sampling and self-calibrated partial parallel reconstruction

Area of Science:

• Medical Imaging
• Biophysics
• Cardiovascular Technology

Background:

• Radial sampling in cardiac MRI offers motion resilience compared to Cartesian methods.
• Undersampling in radial MRI leads to streaking artifacts, limiting acceleration capabilities.

Purpose of the Study:

• To develop and validate a self-calibrated reconstruction technique for suppressing streaking artifacts in highly accelerated parallel radial cardiac MRI.
• To enhance image quality in accelerated cardiac MRI without compromising speed.

Main Methods:

• A novel self-calibrated reconstruction method was developed for 2D and 3D radial k-space data.
• Data were acquired from phantoms and healthy volunteers.
• Image quality was assessed by comparing the proposed method with conventional regridding using statistical analysis and a four-point scoring system.

Main Results:

• The proposed method effectively suppressed undersampling-induced streaking artifacts.
• Significant improvements in image quality were observed (P<.05).
• Image scores improved from 2.14 to 3.34 (undersampling factor 4) and 1.09 to 2.5 (undersampling factor 8).

Conclusions:

• The developed self-calibrated reconstruction method is effective for highly accelerated cardiac MRI.
• It enables improved image quality in parallel radial acquisitions without requiring prior calibration data.
• This technique holds promise for advancing accelerated cardiac imaging applications.