Inversion recovery with embedded self-calibration (IRES).

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Medical Physics
• Biomedical Engineering

Background:

• Parallel MRI acquisition accelerates imaging by using undersampled data and coil sensitivity information.
• Self-calibrated parallel MRI integrates calibration into the scan, avoiding separate calibration scans but reducing net acceleration.
• Existing methods face challenges with reduced acceleration at higher acceleration factors.

Purpose of the Study:

• To develop a novel method for 3D inversion recovery gradient-echo imaging that incorporates self-calibration without compromising the net acceleration factor.
• To evaluate the performance of this new method against standard self-calibration techniques.

Main Methods:

• A new technique, inversion recovery with embedded self-calibration (IRES), was developed for 3D inversion recovery gradient-echo imaging.
• Calibration data were acquired using small tip angles (
• The method was validated using simulations, phantom studies, and in vivo experiments at 3 Tesla.
• Both image-space and k-space based parallel reconstruction methods were employed.

Main Results:

• The IRES method successfully retained effective acceleration at nominal factors of 3 and 4, with comparable Signal-to-Noise Ratio (SNR) and contrast to standard self-calibration.
• At a net 2D acceleration factor of 4, IRES achieved higher SNR than standard self-calibration (nominal acceleration factor of 6) within the same acquisition time.
• The technique demonstrated feasibility across simulation, phantom, and in vivo datasets.

Conclusions:

• The developed IRES method enables efficient self-calibration in 3D inversion recovery MRI without loss of net acceleration.
• IRES offers a promising approach to improve imaging speed and/or SNR in accelerated parallel MRI.
• This technique maintains image quality comparable to conventional methods while enhancing acceleration efficiency.