Scanning problems or ‘how do I find the correct pair of µA and kV?’

When students get to know the handling and technique of the µCT for the first time they get an „aha effect“ followed by thinking about how to keep in mind all the different steps for starting the machine.

First of all, it’s not that complicated after the specimen preparation is done. There are just three options to influence the scan as soon as the position of the object is aligned. X-rays’ movement is not linear, but sinuous. Therefore, source, object, and detector must be as close together as possible, because after crossing the object, the X-rays should not underlie interferences. When the position is determined the setup has to be done.

Goal is to gain as much grey values as possible within the scan area, but without inclosing overexposed white and total black particles. This grey value is made up of the following:

Let’s assume that ten X-ray particles start flying from the source, they compose the current absolute white. When striking the object more particles are absorbed at bigger body parts, and less at thinner parts of the animal. Then a reduced number of particles reaches the detector behind the object and therefore generates a new grey point on the pixel of the detector. Those different grey values can be seen on my histogram. I tried to illustrate it schematically:

20150515_205710 20150515_205720

The more the histogram is used during a scan the better is the resolution in several grey values. We even want to separate two particles from each other with a very similar (not the same) density, so we need to get the smallest different grey values out of it.

Here are two examples: on the left: under-utilised histogram; on the right: almost perfect exploited histogram

 

20150515_200354_LLS   20150515_200717_LLS


Manipulation 1: the timing

It’s the same as with a common camera: the longer the timing, the more particles are absorbed by the detector. The same difference can be seen between a photo made by day and one made by night → at night I need a higher exposure value, otherwise my pics would be too dark or even black. The timing we need for a µCT scan depends on our object:

1.    I have a normal recent animal and would like to scan its bones: with an exposure time of 750ms … an usual value for our machine (http://www.ge-mcs.com/de/radiography-x-ray/ct-computed-tomography/nanotom-s.html)
2.    I have a normal recent animal and would like to scan its chondral: with 1500ms … that’s a longer time, because chondral structures have a lower density than bones
3.    I have a fossil and would like to scan through the stone surrounding it: with 1000ms … which is quite short, because we also use higher kV and µA values
4.    I have an animal stained with Lugol’s Iodine and I’d like to scan the musculature: 1000ms …  because here we use lower kV values
5.    I have a dried insect and would like to scan the interior: 4000ms … because the very fine structures are behind a chitin exoskeleton
6.    I have a dried insect stained with Lugol’s Iodine and I’d like to scan the inner body: with around 1500ms… Staining emphasises the fine interior structures of insects and, therefore, they are clearly distinguishable from the chitin of the cuticula, which doesn’t absorb the staining solution. Thus, we can reduce the timing.

… and so on and so on …

As you can see, everything depends largely on the object!

Of course, values of kV and µA are important as well… so let’s talk about this:

 


Manipulation 2 and 3: kV and µA

Both settings are closely interwoven with each other and are introduced in one go.
The following example is highly abstracted (physicists may forgive me) and shows how both values are defined: kV constitutes the power of a particle; the number of flying particles is called µA.

I often use the following explanation for my students:

In case you want an effected performance: in the field of electrotechnology this means 1 watt. Simplified we could say we want to drag a sofa onto the 5th floor. Well, you have two options now. Either you know strong men, then you need just two of them to lift the sofa, or you’ve bad luck and know only weak men. In this case you need more than two, let’s say six of them…

In science the power of men stands for kV and the number of men is µA. The outcome of this is: 1W = kV*µA.

When you’re thinking now: ‘Ok I’ll make sure everything works so I take six strong men.’, then you’re going to realise quickly that they will get in the way of each other! The same happens here: if kV is set too high while there is too much of µA the X-rays disrupt each other resulting in artefacts on the scan. These are called ring artefacts. We don’t want to work for nothing so we try to avoid these interferences (see picture).

 

Dicreaeosaurus Zahn Test 2 1

For this reason there’s a golden rule: as less as possible, as much as necessary!


noch halb so lang.

2. Ihr wollt ein Fossil scannen:

Ihr stellt also den Stein ins CT und…. seht wieder nichts: Die Grundeinstellungen sind zu niedrig um durch den Stein durchzudringen. . . also was tun? . . . richtig: hoch mit den kV! Eure Röntgenstrahlen müssen stark genug sein, damit wenigstens noch 10% hinten ankommen. Dann kommt ihr aus der undetektierbaren ‘Schwarzen Zone’ heraus und ihr erhaltet wieder Grauwerte und nicht die bösen schwarzen Stellen. Doch nun müsst ihr mit den µA aufpassen! Habt ihr zu viele Teilchen behindern sie sich vor und vor allem auch nach dem Objekt gegenseitig und ihr erhaltet die angedeuteten Ringartefakte! Also lieber: Timing etwas angleichen. . . Ihr werdet feststellen: viel angleichen müsst ihr meistens nicht. Es liegt in den meisten Fällen um 1000ms, weshalb die Scans hier 1,5-2 Stunden dauern können. Bei sehr leichtem Material, wie Kalkstein zum Beispiel, liegen die Scans auch mal bei ca. 1h.


Now as you know the basic principles you surely understand that everything is interdependent!

Let’s have two extreme examples:

1.    You’re going to scan an unstained, dry insect:

Primarily, you don’t see anything on the screen. The animal is overexposed due to basic settings. So what do you do next when you want to scan the structures of the chitin and muscle fibre of a veeery tiny specimen? … Correct: turn down the kV! Otherwise, the particles would just break through the wafer-thin structures instead of being absorbed. This would result in a white picture. With the low kV value you can use a high µA. As high as possible is probably not enough, therefore, you also need a longer timing, often the maximum. … With these settings one scan needs round about 9-12 hours… This is a long time and it’s hard for the cathode and the specimen. For this reason we began to stain insects with iodine. The results are great and it only takes half of the time.

2.    You are going to scan a fossil:

You put a stone into the CT and… again can’t see anything: the basic settings are too low to cross the stone… so what is to do? Correct: up with the kV! There must be enough powerful X-rays so that at least 10% of them reach the back of the chamber. Then you get grey values and no longer the bad black areas. You have to be careful with the µA, though! As you remember, if there are too many particles, they disrupt each other in front and especially behind the specimen. Then you get ring-artefacts again! You better set the correct timing and then you’ll see that you don’t need to adjust a lot more. A proper timing is around 1000ms so that the scans take about 1.5 – 2 hours. Short scans with ca. 1 hour length are possible when light material like chalk is used.

So you see it’s all about trial and error! Don’t be shy, start the action! ^_^
I’m always open for suggestions and questions: write me via mikroctlabor(at)mfn-berlin.de!