磁共振技术--梯度线圈的设计 教学视频文字整理Eng

give a general overview about gradient coils, then talk about gradient coil design, and
finally show some implementations how gradient coils are manufactured.
So to give a broader overview, what is in general necessary for acquiring an NMR signal?
So first we of course need a sample that gives us our signal, so our signal source.
As we’ve heard, we need a magnet to polarize our sample, and the magnet has a strength
of multiple Tesla, and it’s static, so there is no variation in the main field.
And as we’ve just heard, it’s really precise in the parts of parts per million, so not
percent, or yeah, it’s in the parts per million regime.
The next thing we need is an RF coil to excite and acquire our NMR signal, and usually there
are also shims, this could be resistive shims, what we are going to hear about in the next
talk, and shims to correct for inorganicities of the main magnet, as we just heard by Tom.
So with this we basically have an NMR spectrometer, and what I’m going to talk about are the gradient
coils which are added for signal localization.
So this is basically the part that enables for imaging, because it’s used…
Can you advance the slides?
So this is basically what enables to acquire or to map our signals.
So with the gradients we make an MRI out of an NMR.
And gradient coils are given, or the strength is given in millitesla per meter, and the
frequencies which are being used here are in the kilohertz regime compared to, depending
on the signal source here, the megahertz of the RF coil.
So why do we call them gradients?
With our gradient coil we generate a spatially varying field, and for mapping our, or for
changing our local L’Amour frequency, we usually use a constant gradient, so a linear variation
over space, so if you take the spatial derivative we get a constant value, and therefore we
call it gradients.
So this is totally not responding here when I’m clicking.
And because we have this constant variation over space, we have a linear relation between
our object and our frequency space.
And because of this linear relation, yeah, we can excite homogeneous slice profiles,
and I don’t know, and this is also what enables for a straightforward Fourier transform, we
have this linear relation between our object space and our frequency space.
There are usually three orthogonal fields used because we are living in a three-dimensional
world of three spatial dimensions, x, y, y, and z, and what should be noted here is that
our fields that we generate with our gradient coils extends beyond the imaging region.
So we can ask ourselves, we just heard that the main magnet has a really, really high
homogeneity, so we can ask ourselves how linear usually are our encoding fields.
So this is an image that was acquired on a regular clinical 3D scanner in a coronal slice
with a field of view of 500 millimeters.
And is there any way to make this work?
So if we crank this up here, we see that the edges up here, yeah, they are not mis