Analyzing the effects of GRAPPA-based acceleration on MREPT conductivity images acquired with multi spin multi echo pulse sequence

Magnetic resonance electrical properties tomography (MREPT) produces high frequency conductivity images of biological tissues. To produce MREPT images, an MRI pulse sequence using either multi spin multi echo (MSME) or multi gradient echo (MGRE) is used. MSME is better than MGRE in terms of MREPT image quality. However, the data acquisition time is very large. To reduce the acquisition time, application of a commonly used scanning acceleration method, generalized autocalibrating partially parallel acquisitions (GRAPPA), is applied.

GRAPPA speeds up a scan by skipping some of the phase encoding lines which reduces the data acquisition time. While these skipped lines can be approximated using the surrounding acquired k-space lines, this introduces artifacts in MR images. MREPT algorithm relies on image reconstruction from phase images. Therefore, applying GRAPPA in MREPT reconstruction requires a new method of combining phase images from multichannel k-space data to calculate transreceive phase and subsequently MREPT conductivity images.

To explore the application of GRAPPA accelerated MSME scans in MREPT, a multi subject pipeline was developed in Python and MATLAB. The errors between GRAPPA accelerated magnitude images and the corresponding full k-space magnitude images were calculated and found within acceptable errors. MREPT was performed on 2D slices covering the whole brain volume of 5 subjects. These slices were then segmented into cerebrospinal fluid, gray matter and white matter using segmentation masks from FSL. The relative conductivity between the tissues was found to be consistent across subjects. The effect of the acceleration factor on contrast among the three tissues was consistent across the different acceleration factors and across different subjects. In conclusion, GRAPPA accelerated MREPT can be performed using MSME pulse sequence to produce high frequency conductivity images with acceptable image quality.