JEMRIS  2.8.1
open-source MRI simulations
Example 2: Position Encoding with Nonlinear Fields

This section provides an example on using nonlinear gradients with JEMRIS.

First, create in Matlab a 1-D sample consisting of a repeated step function:

for i=1:nb
M0=c; T1=100*c; T2=100*c; T2s=T2; DB=0*c;
RES=1; OFFSET=[0 0 0];
save blocked1D M0 T1 T2 T2s DB RES OFFSET

Then, load the mat-file "blocked1D" from the simulation GUI ("user defined sample").

Next, a simple 1D gradient echo imaging sequence looks like this:

<?xml version="1.0" encoding="utf-8"?>
<Parameters FOVx="1024" FOVy="1" FOVz="1" Name="P" Nx="1024" Ny="1" Nz="1" TD="25" TE="5" TR="500">
<ConcatSequence Name="C">
<HARDRFPULSE Axis="RF" Duration="0.1" FlipAngle="90" Name="P1" Observe="C,Counter/C,Repetitions/P,FOVx"/>
<TRAPGRADPULSE Axis="GX" FlatTopArea="-a1/2" FlatTopTime="4" Name="P2" Observe="P3,Area"/>
<TRAPGRADPULSE ADCs="a2" Axis="GX" FlatTopArea="2*a1" FlatTopTime="8" Name="P3" Observe="P,KMAXx/P,Nx"/>

Load this sequence into the simulation GUI and run it. The 1-D image shows the expected result.

Now open the sequence with the sequence GUI and change the "NLG_field" attribute for both gradient pulses (P2 and P3) to


in order to perform cubic spatial encoding.

Alternatively, the formulas


uses the arctangent function, and


a Rayleigh function (resembling the PATLOC approach).

Below the result for the arctangent is depicted. The 1-D image shows the higher resolution encoding in the middle of the sample where the arctangent function has the largest slope, i.e. the strongest "gradient" encoding takes place.

These simulations only need some seconds on a single CPU.


-- last change 17.06.2016 | Tony Stoecker | Imprint --