High speed X-ray video: jumping beans, wind-up toys and more!

today on applied science I'd like to show you some x-ray video but not just any x-ray video high speed x-ray video captured with this Photon counting camera that can do about 300 frames a second so let me show you some of the footage that I acquired and then we'll talk about how the camera works finding subjects for this setup is actually pretty tricky because the thing needs to pass enough x-rays to make a good image so anything with a metal plate like a pocket watch isn't going to work the best subjects are made of water or plastic and allow enough x-rays to get through to make an image I was especially excited about these These are Mexican jumping beans something for my childhood and you always wonder what's inside there I mean you know it's an insect larva but exactly what's going on um they do jump quite well these are the most active jumping beans that I've ever seen put a link in the description keep track of this guy on the left yeah I got some real air time also making good use of the high speed things that are Dynamic you know fluids that move around are things that bounce or jump or interesting subjects for high speed video in this case I turned the frame rate back down to about 60 frames a second because getting again getting enough x-rays even through this plain old switch was a problem I think I cranked the frame rate back up to about two or three hundred for this one this one is small enough where you can get a good image and you can see it's a little bit noisier that the household light switch was less noisy because the frame rate was lower but here you can actually see the mechanism inside there and see my finger bones too now this is interesting this is actually a knockoff micro switch not a name brand and when you see we look in there it's actually just pushing sideways on that coil spring which is probably not a great design for longevity in contrast this is a commercial name brand micro switch rated for a high number of cycles and you can see internally it's built completely differently there's no coil spring it's a leaf spring and it's even like double articulated so it's it has a very controlled movement inside there I also got the feeling that you know living creatures would be an interesting thing to sub to test and insects are a great place to start because they're kind of everywhere and small and this insect in particular a salad bug has a shell which is has a different radio transparency than the rest of it and so you get a really good high resolution image of them and then I thought you know really the best thing would be to get a cricket jumping like that would be the best use of this high-speed x-ray setup and so I got some crickets from the store and spent many days trying to get them to jump on command and never got it to happen actually surprisingly they were jumping in the car but not jumping here but I was still thinking of jumping and so why not plastic and so they have these little you know plastic poppers another toy from the childhood you know at least I got that on camera jumping and then finally I was looking for some um fluid dynamics that would be interesting and so what we're doing here is dripping uh potassium iodide solution into water and not surprisingly it looks a lot like dripping food coloring in but it's pretty cool to think that the contrast there is coming from additional x-ray absorption just from that potassium iodide dissolved in in the stuff that I'm dripping in there's another good wind-up toy it's kind of jumping around so you see the Dynamics there and in this case you know it was weird that stood out to me on this one is that there's some kind of contamination under the toys fur like you see all those black spots there there's some kind of high molecular weight or metal or something that's mixed in with the toys fur under the fur or something I'm gonna have to get the xrf gun out and find out what this thing is contaminated with funny and then you know any insect that happened to fly through the shop I caught and you know just wanted to see how it was going to look in the thing one interesting thing is that they all have this air sac inside them these insects this is the last insect um a butterfly that I found in the garden and the detail that you can see on this is pretty cool you can see the coiled up mouth part there you can even see the butterflies wings with pretty good detail it's impressive that you know something as thin as a butterfly wing catches enough x-rays to be visible there in order to capture these high-speed X-Ray videos the sensor has to be extremely sensitive capturing as much x-ray energy as it possibly can and turning it into an image and it does this by Photon counting which is a fundamentally different way of operating than a standard camera sensor so in a previous video I had a large area x-ray detector and it works basically by you know it's very similar to a large format sensor that you would find in a normal digital camera and internally it has a phosphorus screen bonded to the sensor and the whole thing is in the dark so when an x-ray comes in it causes the phosphorus screen to produce a little flash of light the light goes into the camera and the way that those camera sensors work is it's basically like a little bucket of storage of energy storage so an incoming Photon knocks an electron into the bucket and then at the end of the exposure period you basically measure how many electrons you have in the bucket using an analog digital converter and this is fine but there's definitely losses throughout this process so for example an incoming x-ray might cause a flash of light from the phosphor screen but then that phosphor uh you know ejected Photon the visible Photon goes in the wrong direction it doesn't go into the image sensor or you start filling up your bucket and there's noise in your analog digital converter there's all these processes so at the end of the day an exposure with that old setup was about five seconds now fair enough I had the tube much further away than I do with the setup here but still a five second exposure would imply a 0.2 frame per second frame rate and this is able to capture video at 300 frames per second so we're talking about a thousand times more exposure and or sensitivity to make up a good image so the way this works is it's it's a direct conversion sensor so in an x-ray Photon goes in there's a semiconductor that's specially tuned to catch x-ray photons between 5 and 30 kiloelectron volts and when the photon goes in there's a chance that it is instantly converted into a digital pulse so it just counts photons and it's cool the output of this is a 32-bit Tiff image and that number is how many photons hit the pixel during your exposure and you can choose to have an exposure from like a microsecond up to a million seconds and it just sits there and counts just counts photons I think the rated dynamic range of this device is 20 bits or about a million to one which is very high extremely high dynamic range as you might have guessed this specialized piece of equipment is extremely expensive and we can thank the good Folks at dextrous for loaning this to me because there's no way I'd be able to afford it but it gets even better dextrous is pushing the state of the art and their latest detectors have a field of view that is substantially larger and can capture eight megapixels at several thousand frames per second and so maybe if we all ask nicely and the sales rep schedule aligns with mine we can capture some video with that when I asked about their latest highest performance detectors the rep said well those are those are substantial and for some reason I got this image in my mind that they send the sales rep out with it in a briefcase kind of handcuffed to his uh you know wrist or something but with either very very specialized very expensive pieces of equipment the care and feeding of this detector are also pretty unusual when I'm not using it I keep it in a bag with this big desiccant pack because it's moisture sensitive and when the sensor is in use it has a port for dry nitrogen or argon which you plug in there and it just blows about half a liter a minute of dry gas through there because it is sensitive to moisture and the front of it is this mirror looking thing it's actually a very thin piece of mylar that I have a feeling if you touch it that might destroy the camera so it has a very very sensitive and specialized and so you might be wondering what do you actually do with these things I get the feeling that most of the clients are actually not interested in capturing high-speed x-ray video like Shadow grams like I was showing in this video These are typically used at beam lines at particle accelerator installations where you might want to look at xrf so you're going to shoot your particle Beam at something and you want to capture the x-rays that are scattered off of that interaction so these are typically set up in a way that the beam doesn't shoot right into the detector it's just looking at the stuff that's scattered from what you're aiming at places like eater and other like giant accelerator you know particle physics institutions would be interested in this detector I get the feeling that Imaging is actually a fairly unusual use for this believe it or not the system also includes this rack mount server which is running a customized version of Linux optimized for high-speed real-time data transfer and the physical link between the computer and the detector is this PCI card called a gigastar which I've never heard of before but remember that these images are uncompressed for a lot of the scientific uses that you'd want to use this system for compressing the image would ruin all that so it's sending a 32-bit per pixel image at these high frame rates into the computer and storing them as uncompressed 32-bit tiffs so that the data rates can be very high a typical acquisition works like this with with everything set up in place I'll turn the X-ray tube on and then run the acquisition command which will start pulling data in from the sensor and after the acquisition is complete you're basically left with hundreds or thousands of these 32-bit Tiff files and as it turns out viewing those Tiff files in a video format is quite difficult so when you're recording these things you got to have sort of a way of determining if you've got a good take or not so setting something up a bouncing ball or a jumping Cricket or whatever it is you gotta you know you have no way of knowing if you've got the actual action or not when the acquisition is over so um opening a 32-bit Tiff is challenging enough this version of Linux comes with a program called image magic which doesn't even open 32-bit tiffs you have to recompile it and I wasn't going to do all that so I found out the best thing to do is to use image J which can open 32-bit tiffs and even batch save them at 16-bit tiffs that I can then import into my non-linear editing software resolve and that actually worked out fine so the whole pipeline actually ended up being okay just figuring it out was was a little bit of a challenge quick note on x-ray safety just like in my last x-ray video I described the precautions that I'm taking so the X-ray tube that I'm using is a 50 kilovolt one milliamp tube which is very comparable to what you would encounter in a dentist's office so the same safety protocols that you as a patient and the dentist would take apply here as well the main thing is to stop the main beam so I have this steel setup in back of where I'm shooting the x-rays so that the primary beam is always intercepted by the heavy steel and I've used a geiger counter to verify that there's not any primary beam that gets through there now there's still Backscatter of course the geiger counter is beeping even though it's not in front of the X-ray tube due to Backscatter and so I've also walked around the entire garage with the geiger counter to verify that the Backscatter doesn't get out through the walls so when I'm using this the entire shop is the enclosure for this x-ray system another random question why does a Mexican jumping bean jump sounds a lot like why did the chicken cross the road but no this is actually a real answer it's trying to get out of the sunlight and I'll show you I didn't realize how light sensitive these things are but you can hear the just listen for the clicking and then when I turn the lights out completely quiet and when you flip the lights back on they're back to jumping around again so when they're really active like this they're trying to jump out of the sunlight to get in the shade so that they don't overheat and as the day goes on you know the sunlight might change where it's shining on the ground so they start jumping around to kind of follow the shadows and stay in the cool so I have a feeling that high speed x-ray video will show up again on my channel maybe with the super high-end detector or this one again finding subjects that are good to video with a small field of view and if this frame rate is pretty challenging and I know some of you are thinking why didn't I just get a mouse or something remember that this is YouTube and if I do that I'm probably going to want to work with a real Kinesiology researcher to to make actual research you know good use of the of the video in that case which I think it would be I think super high speed video of a mouse moving would be pretty pretty useful and pretty cool to watch well anyway I hope you found that interesting see you next time bye

Channel: Applied Science

Views: 246,744

Keywords:

Id: xdpDd7dyU00

Length: 13min 9sec (789 seconds)

Published: Mon Sep 12 2022