MRI: parallel MRI reconstruction
These Matlab routines include Matlab functions simulating and reconstructing data acquired from a multi-channel coil array using SENSE or GRAPPA algorithms.
Here are examples of the un-accelerated MRI, 2X accelerated SENSE and GRAPPA reconstructions, 3X SENSE accelerated SENSE reconstruction and 4X GRAPPA reconstruction. 
Please download codes here
1D GRAPPA reconstruction: GRAPPA reconstruction uses data from an 8-channel head array coil. The data were simulated from the combination of B1 fields by Biot-Savart’s law and anatomical features from MPRAGE sequences. This example simulated R-fold accelerations (R=2, 3, or 4) in A/P phase-encoding direction. With image size = 128 x 128, 32 auto-calibration scan (ACS) lines are used to estimate the GRAPPA reconstruction coefficients with a 4-by-3 block, which means 4 (+/-2) phase encoding data and 3 (-1, 0, +1) frequency encoding data were used to fit ACS lines.
Code: grappa_8ch_example.m
1D SENSE reconstruction: SENSE reconstruction uses data from an 8-channel head array coil . The data were simulated from the combination of B1 fields by Biot-Savart’s law and anatomical features from MPRAGE sequences. Both SENSE reconstructed images and the g-factor maps are calculated.
Code: sense_8ch_example.m