 Thank you. I'm not sure quite how I follow the uses and abuses of telly dildonics and internet things But I'll show you uses and uses of CT scan a little bit of work. We've been doing Show you some of the stuff we've actually been doing to recover some lost film and just some lost cultural heritage So the first question that comes up generally Do people know the difference between a CT scanner and an MRI machine? Well, they're both big things you lay down in They both make a noise CT bombards you with ionizing radiation and MRI makes your protons vibrate I Build and abuse CT scanners. I want to build a small desktop MRI, but I haven't got around to it yet So this is a medical CT Useful horses. I couldn't get permission. It's really difficult to get ethics approval to show your patient going into a CT But no one seems to care too much about horses So the big donut shape is actually a rotating x-ray source with a detector opposite it The sort of CT scanners I use are more research systems So desktop lab scale systems Where the scanning volume is not a horse's head. It's approximately a Pringles can so if it fits in a Pringles can it can go my CT scanner and A Pringles can in this particular scanner would take around about a week to scan It's not a it's not a fast scanner, but it goes for ultimate contrast You can see things you just won't see in commercial systems This was built specifically for some dental research. So all the stuff. I'm showing the abuses of the CT scanners No patients were stopped from getting their surgery or their treatment due to what I've been doing with CT This is all research machinery So I said we use ionizing radiation Where where am I operating? I'm operating up in the x-ray region of the spectrum So past the visible spectrum through the ultraviolet into the x-rays Generally round about 90 kV x-ray energy Which is not actually a scale shown on that spectrum, but Closer to the UV end of the x-ray spectrum It's a hard enough x-rays to go through your bone And so just take your step back from CT. What do you do if you just invented x-rays? So you're like these people You've connected a high voltage x-ray source, which is the the black tube in the background to a vacuum Source vacuum bulb Which is some of the things that are on the back shelf In the picture and the guy is actually looking at his hand through a phosphorescent screen The two of them are both imaging their hands So if you just invented x-rays you look at your hands if you're the guy that actually invented x-rays or first discovered Wilhelm Rontgen, what you say is that Bertha darling, please put your hand in this dangerous looking apparatus for me while I go outside the room And you get the very first X-ray image medical x-ray image taken at Bertha Rontgen's hand Showing the bones and showing her gold wedding ring She is recorded as saying I have seen my death because she's seen how she's going to end up as a skeleton She didn't die from hand cancer or anything. She she lived to a fairly ripe old age for the 1860s But x-rays are boring You know if you've ever had an x-ray you've you've generally had an x-ray as you've done something like this compound fracture of the arm You're having a bad day if you're having an x-ray But I don't do x-rays I do CT scans and if you're having a CT scan a medical CT scan CT scans are terrifying Because you're having a CT scan if you've had a really bad day You've gone through a car windscreen at speed or you're diagnosed with a tumour in this particular case This is a simulated image So you really only have CT when there's something seriously wrong But I don't do medical CT so just to reiterate there's no patients were harmed in the the messing around I've been doing So what do I do it an actual day job the use of CT scanners I'd scan lots of teeth I work in a dental school So I'd scan lots of extracted human teeth that are going to be used for research projects Or when the boss comes in and says my seven-year-old daughter has just lost one of her teeth She wants to see what it looks like inside that goes in the scanner And you wonder why seven-year-olds have teeth that look like this and have that many fillings and I'm not going to comment I also scan lots of bones. This was actually an extracted femoral head from one of our technicians She had to have her hip replaced getting quite elderly and She was getting hip pain what we actually see are these protrusions coming up at the surface These were scraping the inside of her The hip joint and causing pain. This is the reason the hip was replaced So this is a human femoral head which we scanned in our scanner took about three days to scan And we are able to these protrusions are force-coloured, but they're actually much harder than the bone There's more mineral density. So with our scanner with the extra contrast we can see this You probably wouldn't really see in a commercial CT or medical CT Other bones I sometimes scan are in much worse shape. This is Collapsed I think it's third and fourth lower vertebra might be second and third That have actually collapsed in this poor guy and fused It didn't have a happy ending for him because I've got a section of his lower spine in my lab So, you know, we got this after been donated to medical research But that's all sort of the day job. That's what I get paid for doing scanning teeth and bones Borderline from the use to the abuse is more looking at things like this. So this is actually an acorn That's got parasitic wasps living inside it These were used in the Middle Ages. Well from early antiquity to the Middle Ages You grind these up Mix them with iron sulfate and you can make a really nice dark black ink Which is a permanent ink and the history of the world is written in the ink made from this It's called iron gall ink And you can actually see in this. So this is this is rendered CT data a volumetric rendered CT data And you can see that there's chambers in there and the wasps are growing in it and this was obviously harvested before they hatched out We did discover something with this the literature we were reading says there's only ever one wasp per acorn in this one We count at least six. We went and found some entomologists on they went. Oh, yeah, the literature is wrong. Everyone knows that No one's got round the correcting it So this led on to well if we can actually see The the oak apple scan or the iron the oak gall scan occurred After this particular piece of work because someone said well if you can see very low levels of contrast in bone or teeth Can you actually see the iron in iron gall ink on paper or on parchment? And it turns out we can we did some initial test scans and we could see and we could read rolled up documents So then the Norfolk Record Office got in contact with us and said well, we've got this 14th century parchment roll It's them report on the manner of pressing them. It's basically the yearly accounts This is as far as it will unroll it's got wet at some point in its history It's written on parchment parchment is processed animal skin when it gets wet and gets compressed it basically turns into a raw hide dog chew If you've ever tried to unpeel a raw hide dog chew, you know, it doesn't work You can see there's tear holes in this where someone's tried to peel it apart and it's just ripped So we scan this It took I think this took about five days to scan And what we were able to do is Produce a full virtual unrolling of this if you look on YouTube Search for the apocalypto project You might want to put minus Mel Gibson in that to just get out we named the project before he named the film You'll find an uploaded Section that the BBC did on the work. We did this and there's some links to some other information But what we were able to show is the top section is invisible light the lower section is just from the x-ray data And we're able to see ink and contrast in places where it's even faded optically. So this was pretty good And We published this and then we got contacted by these people and these people have a really cool project They've got this chest of a 2,600 undilivered 17th century letters of which approximately 600 have never been opened Have a look at the Brienne project. This is ongoing if you're in the Netherlands In the Hague go to the Postal Museum on Museum of Communication. You can see this chest You can see the work we're doing with this What they wanted to do was Read well what the the historians that are interested in this project They want to know what the unopened letters say they can read the ones that are open But what's in the the ones that have not been opened? So we got involved and they have these wonderful wax seals So they don't want to open the ones that are not yet opened for many reasons that they go on to on the Brienne project website But effectively you can't study a locked 17th century letter if you've opened it. It's no longer a locked letter And they have all these wonderful seals and they form their own security That should be a really good EMF camp workshop somebody to do the letter locking for that I'll suggest it for next time So we scan some of the letters we stack them up This is a sort of Swiss roll section so the letters are projecting out of the board towards you This is a section through a slice through roughly halfway You can see these are pretty complicated. There's many layers. There's lots of bright spots The bright spots in this are actually crushed a seashell that's been used to absorb some of the ink to dry the ink on the page The ink shows up as much fainter bright spots But if you remember those lovely 17th century wax seals, they're full of lead You don't get many x-rays through the lead and the reconstruction algorithm We use just cannot cope until you get all these white areas and streaking and failure Working with a couple of really bright students at MIT at the moment to Try to do the unfolding of these and do digital origami to virtually unfold these letters make them readable We're hoping to publish this year So after we've done all of this people were keen to see what else we could do and We got sent this photo and It looks like cine film and It came along with an email that said Effectively this it's triacetate film It degrades it releases acetic acid. This is triacetate film is basically polymerized vinegar. It turns back into vinegar The problem is it's also auto catalytic So when the vinegar is formed on the acetic acid is formed it forms more It helps degrade faster. It forms more of the acetic acid helps do the helps degrade faster The emulsion that's pasted on the actual part of the film that contains the image Then becomes cracked distorted. It slides off of the acetate base It's no longer usable And then the layers of film as this progresses further the layers of the film actually get stuck together So that if you actually try to unroll with this the way it was described to us was if you take a sheet of glass Draw an image of a stick figure in Marmite or ketchup put another sheet of glass on the top It'll be recognizable, but as soon as you try to peel those apart the image just smears So what usually happens in archives is as soon as there's the first hint of vinegar about a film it immediately gets put through a Capture system to get what they can from it a Recording is made the original goes in the bin They don't want this because if there's a acetic acid fumes in the archive it can infect other film It can just it can cause real problems So and usually there's more than one copy of a particular film So we we did approve the concept we kept putting the guy off because one This is bigger than a Pringles can this won't fit in my scanner So we we and we also didn't think it was gonna work But the guy from the BBC Charles Norton kept on and he sent us this For the younger people in the audience that don't recognize what this is It's sort of the prototype micro SD card of its day Thank you So this is actually a 16 millimeter positive negative Sorry positive print of Some film the Garmott British news with a soundtrack it fits in the 35 mil film can which fits in my scanner quite nicely We put it in the scanner left it for a weekend and expected to come back in Monday just to nothing we reconstructed the image and Just paging through in some very simple software You can see actually we've got pretty good separation of the layers of the film in the Swiss roll slice And if you just cut it at any particular angle through Ah, there's some pictures in this we might actually have to do this We were really expecting to come back to an empty data set where nothing's really visible And it turns out we do have the contrast sensitivity to see silver Deposited silver on tracetate stock So this caused a bit of a problem because we didn't expect to have to do this We didn't expect it will work So we ended up spending The plan was to spend maybe an afternoon. This will be a quick hack We'll see if we can get anything out of this at the end of a quick afternoon. It was like ah, this this does actually work We'll spend another day on it So the guy I work with grand Davis who is an absolute genius he spent most of the time this he did the film I was working on the soundtrack He spent a week on this it was always like I just need another afternoon I can just tweak this I can just tweak this and we ended up with this so we came in on the Monday The sound's not playing ah So there is actually a soundtrack for that. It didn't play. Let me just that might be my fault. I Have unmuted it. No, I've just got nowhere getting sound out of this at the moment It basically plays a soundtrack and the soundtrack was the bit I worked on on this the soundtrack in cine film Was just recorded as an amplitude modulation wavy line Along the side Okay, let's try this with sound. I've heard this so many times, but I like this Well, let's go back German So it's a bit crackly, but the sample rate is limited by the resolution we scan that Yeah, it's a real physical limit We scan this at about a 30 micron voxel resolution which limits the playback audio bandwidth to about 7 kilohertz So after we've done this and we actually received this we sent this to Charles And he was like will you please have a go with our film and he told us a bit more about what it was It's this it's more common wires series one episode two It was found in a shed in Nigeria It's print from telly Cine The film can it was actually on the bottom of the stack. So the film can has got cracked It basically means the film inside has been exposed to the elements for six fifty sixty years And it's in a pretty degraded state What used to happen is the BBC took a load of stuff on videotape But videotape is expensive to want you to reuse that series and there's a lot missing doctor who and a lot of lost more Come a wise and a lot of early TV shows the stuff that survived was the stuff that got telly print or telly Cine And sent out to South Africa, New Zealand Australia on the understanding they would you burn after showing but this one survived Um It's too big to fit in my scanner. It does not fit In a Pringles can it does fit in a laser cutter It turned out if you really want to scan this at the resolution we think we need you need to cut it into roughly one inch cubes You could you can't put this through a projector it's already it would never go through a projector So they said what the hell got nothing to lose So we did the first scan again. It took roughly a weekend of a one-inch cube. We were paging through The the scan data and we saw this image appeared so roughly the middle of that So people of a certain age and I count myself in that you immediately recognize Eric Malcolm The face kind of stands out especially when you're you're at the monitor We're looking at it in work is almost the size of that and we're sitting up close to it You recognize Eric Malcolm. So it was like, oh, this does actually work The software that had been written to played a gamut British news didn't work for this because that was relied on a continual Spyrocatrace. This is a one-inch cube with lots of disjoint edges So Graham worked on it. He got some pretty good software working and then The BBC got involved. So this guy had a BBC R&D whose day job is to do all the really cool fun stuff Took what Graham and myself had worked on and wrote some other software So he did a lot of statistical methods to just try to identify what the acetate layer is because it merged you get Bridging layers of the goo of the actual film layer bridges the acetate He did a lot of stats on this I was able to even you know account for the warping because these are warped layers and start extracting full frames And the software was written. I think this is MATLAB But it would it gave a lot of options loading the data look at the actual layers which are shown on the left-hand side and then Attempt to process this as close to automatic as possible the software we wrote at Queen Mary was very manual This this was aimed to be automatic And it produces some really nice images It takes about three to six minutes per layer So each block each one inch cube is roughly six to nine hours But you don't get complete Detection near the edges. Well part of that is when you laser cut the film You're not necessarily cutting on an image boundary you we generally get one and a half to two images in a block And those are not always complete images So the good thing is nothing much changes between frames if you get half at the top of a good image and the bottom of another one You can merge them And so Adam worked very hard on this And some of this was actually published on the BBC R&D blog around Christmas time this year So my phone started going mad lots of people. I've just seen this on Twitter. This is what you've been working on, isn't it? Christmas So to give some idea we ended up we scanned About 35 blocks the 35 sections the film is degrading the whole time each one takes a weekend to scan We do have some other commitments for things we needed to scan so we couldn't be always scanning film So we ended up we did a complete circuit of the outside of the reel into one inch cubes And then we just cut a radial section straight through and scan those because as you're getting closer to the center of the reel The layers are getting closer together the time periods not not such a large time period between layers And we produced this sort of thing Then Adam had to take to the next level. Oh The other thing is an audio recording of this episode existed It was the only thing that they knew existed some guy in Australia made a reel-to-reel tape recording of this from the TV So this is what we have so far Miming to Bing Crosby's voice and beautifully as well Not a word out of sync. No, I'm not miming now. It's me was You realize of course that the tape has stopped How does he do it it's a thrill working with a small little genius It's me who's talking now. I'm not talking like me There we unfortunately fade out. Um, so this is still very much work in progress We're hoping to get this cleaned up and hopefully get the whole tea the whole thing will be on the TV Hopefully for one Christmas because it's more common wise. It's not Christmas. We're at more common wise um that is sort of Some of the abuse of CT scanners this is where we've got to so We're getting near for actually producing a whole TV episode from stuff that we didn't think would ever work It had no reason to ever work, but you know, our kit is better than we think One other thing just a quick aside total abuse of CT scanners This is my lab. I sit at that desk You know a sizable portion of the day near a CT scanner processing CT results Well, if you've got this and you can scan what you like, of course you do One day I had an apple for lunch. There was a hole in it I didn't realize that so I was about to take a bite out of it Well, you put it in the CT scanner right put it just use it as a radiography machine Just use it take an x-ray am I going to see the maggot? No, I put it in there was no maggot in my apple So I posted this up the Twitter hashtag x-ray my lunch and I put a couple others in This was another one going back to my East End roots any guesses Jolly deals But of course the day you don't see see the apple you take in for lunch and you take a bite out of it This is what happens Just one other very quick thing because this is like we did for one of our students at work Any guesses what this is? Yeah, so an iPhone 3 iPhone 4 Um, but the iPhone was running when we put it in there So what we wondered was what's the camera of an iPhone C as it goes past the x-ray emitter? First thing is it screws up the back lighting and stuff. You saw these white dots. They're not compression artifacts They're x-ray photo was hitting the detector I'm surprised it lasted as long as it did We've actually been now trying this phone's been donated to us We're trying to kill it in the scanner and it just isn't dead yet Uh, just one other thing a friend of mine. We do smart. We bought some pound shop USB hubs and they didn't work And then we took the back off And if you actually look at the pins for these USB ports you're all in parallel That doesn't work What's under the black blob then? nothing There's bond out pads they all Someone must have sat down and done this as a practice, you know for how to lay out a PCB and there's okay soldiers and bits on the cell So yeah, that's it. That's the end of my talk for now, but uh, thank you Thank you very much David. I think the plan was that David would Take questions in the bar or somewhere like that So we won't do questions here, but thank you all very much for listening