0:43

So to fill the multiple k-spaces for the multiple slices,

there are two different ways to fill them.

So one is to fill all k-space lines, in one slice and then move to next slices.

And as shown in this figure, we may have a multiple slices as shown here,

and then we can acquire the one k-space line, one with outline,

and then we can move to next k-space line for the same slice.

And then we can repeat this procedure to fill

all k-space lines in the slice number one.

And we can move to next k-space line for k-space line for the next slice and

then we can fill all of them and then move to the next slice.

And this can be one possibility.

And in this case, scan time will be number of phase encoding

lines here multiplied by number of slices, this one here,

and then that is multiplied by times repeated for each step.

And then that is multiplied by a number of averaging.

So the number of averaging is performed in my location to improve similar

to noise ratio, for which it is similar to other data location strategies,

like communication systems or similar systems.

So we perform the same data location multiple times,

then that improves signal to noise ratio.

So in proportion to square root, number of averaging.

So signal increase by a factor of N, and

noise increases by a factor of square root N.

So overall increase by a scale to n.

So this is going to be all our scan time for this data location.

And then in order to save the scan time short TR is required

because we have to acquire data multiple times for each slice and

then to fill the case base multiple times.

So each preferred in some gradient echo applications with fast acquisition,

but it is not that common.

There are some special application that requires this acquisition mode.

But the more common are data acquisition strategies we want to

be just mentioned here.

So another way to fill the same phase-encoding line across

the slices first and then move to next phase-encoding lines.

So as shown in this figure, we can fill one k-space line for

the slice number one and then move to next slice, slice number two,

we may fill the same k-space position in the slice number two.

And then we'll do slice number three, okay.

And then, if you fill all of them, for

the across slices at the same phase encoding lines, and

then we may move to next phase encoding line for the different slices.

And then this is going to be repeated across slices.

So in this case time to repeat is not from here to here, but

from here to here, and as I mentioned in some previous weeks,

this time to repeat the [INAUDIBLE] is about, exciting the same position, okay?

So this time to repeat is just slightly longer because to

accommodate data location for each line, across slices.

So the time to repeat is going to be a little bit longer.

So in this case, scan time is not depending on this procedure,

scan time is depending on this whole TR.

Because of that the scan time will be the number of phase encoding line.

So that is going to be repeated,

this TR is going to be repeated number of phasing coding lines times.

So N phasing coding line multiplied by TR, and

that is multiplied by number of averaging.

So this is going be the scan time.

So here the scan time does not depend on the number of slices, okay?

And TR should be longer, so

TR should be long enough to accommodate multiple slices in one TR.

And imaging with long TR that increases a signal-to-noise ratio.

So, if this time from here to here is the same as the previous occasion mode,

the total scan time will remain the same.

But this time to repeat increased much longer than the previous case.

So a signal is going to increase compared to the previous occasion model.

So this scheme is preferred because of this,

this scheme is preferred in most clinical 2D imaging cases like T1 imaging,

T2 imaging, and T2 star weighted imaging or Proton Density weighted imaging.

5:47

Okay, let's talk about three dimensional imaging.

So we talked about in most cases in two dimensional multi slice imaging,

we just talked about in previous two slices.

And this is occasion mode of two dimensional-multislice imaging,

and this is going to be three dimensional polymetric imaging.

So for the 2D slice, we have to apply for slice selection and slice refocusing,

and phasing encoding and read out prephasing, and read out.

So this is going to be repeated across the slices, and

also across phase encoding lines.

And then in case of three dimensional imaging the pause sequence is similar,

but slightly different from the previous two dimensional case.

So as shown here, we may still apply for some gradient,

in this case selection gradient.

So the gradient select this whole three-dimensional volume, okay,

whole three-dimensional volume is excited.

So because of that, this gradient strength is going to be

much smaller than the two-dimensional slice selection case.

And then because this gradient is reduced,

slice refocusing gradient will be also reduced.

But to distinguish spatial information along this selection direction,

we have to apply for a second phase, including gradient, as shown here.

And there are still first phase-encoding to distinguish in plane spatial

information.

And then we have read out prephasing gradient and

read out gradient, other things are very similar.

So the only difference here is second phase encoding gradient,

which is applied along this slab selection direction, okay.

So this phase encoding gradient changing means this occasion procedure is

repeated by changing this gradient strength and also this gradient strength.

So phase-encoding gradient is distinctive a lot,

they use two dimensionally so

that means the scan time increases a lot, okay?

So because of that the three dimensional imaging needed to the TR for

this three dimensional imaging should be shorter,okay.

So otherwise the scan time is going to be pretty long.

But there are some advantages for three dimensional imaging too.

We'll talk about them later.

For these two dimensional imaging and three dimensional imaging cases,

we can compare how they fill in each case, how data filled.

So for the two dimensional case, we have a K-space for splice number 1,

2, and N, and N + 1.

So each of them separated, so the 2D k-spaces are separately acquired.

And then two dimensional view of transfer is applied to each

slices to get difference slices, as shown here.

Well, in case of three dimensional imaging,

there are only one single three dimensional K-space, okay?

One single three dimensional K-space and we have to apply for

three dimensional fusion transform.

Now, slab selection direction is not separated because whole 3D

volume is excited all together, but they are separated in the viewpoint

of phase encoding applied along slab selection direction.

So there are multiple k-spaces, but they are part of the one big 3D k-space.

So we have to apply for fewer along this slab selection direction too, okay.

So we have to apply for three dimensional fourier transform.

And then we can get a similar volumetric data, similar to the two dimensional case.

But in this case, three dimensional case, there's no gap between the slice.

So that is the slice different from the two dimensional case.

In case of two dimension case we can apply for gap between the slices, but

in case of three dimension there's no gap between the slices.

9:58

Okay, so the advantage of three dimensional imaging is that the high

spatial resolution is possible along the slice direction, okay.

So in case of two dimensional imaging,

slight synchronization very slim then the signal is very poor, okay.

But in case of three dimensional imaging,

whole volume is excited all together so they gives relatively higher

signal to noise ratio because of volume averaging effect.

So it's about exciting the whole volume in three dimensional, okay?

Which is much bigger than two-dimensional imaging.

We probably want to do very thin slice imaging,

then three-dimensional imaging is very advantageous.

10:43

So the high spatial resolution along slice direction is now possible for

three-dimensional imaging.

So it provide high sensitivity for detecting small lesion, okay?

Three dimensional gradient echo imaging is extremely short TR can be used, okay?

After magnetization preparation to enhance contrast.

So, whole selected volume can be excited all together, but with a very short TR and

T and there will not be very much contrast but scan time is pretty fast.

But that kind of data location mode can be combined with

magnetization preparation to give special contrast that we desire, okay?

Sometimes in budget recovery can be applied to enhance T1 contrast or

some physiological imaging can be used, okay?

So this is advantage of three dimensional imaging.

But this disadvantages over three dimensional imaging is that because

short TR is necessary to maintain clinical revision with scan time.

Though scan time it can increase if we use a regular TR that is used for

two dimension imaging.

The reason is, we have to repeat the location by changing not just the first

phase encoding direction, but also the slice selection also need additional

phase encoding, which should be repeated, that is multiplied.

So scan time increases significantly, but to maintain that,

we have to use a short TR so that is a kind of requirement for the three

dimensional imaging that limit the desired contrast for certain imaging method.

So, in many cases, the three dimensional imaging can be used as is, but also it can

be combined with special magnetization preparation to enhance dimension.

So, as I mentioned here, so,

in three dimensional imaging the short TR decreases SNR,

but this can be compensated by the whole volume averaging effect, okay?

So these two factors compensate each other.