Corpse pose
October 29, 2014 10:09 PM Subscribe
Did anyone else but me wonder if this would have been a fatal dose of radiation if it had used real x-rays instead of CG?
posted by double block and bleed at 4:18 AM on October 30, 2014
posted by double block and bleed at 4:18 AM on October 30, 2014
Cruising through the spine is kind of cool. But I think if they had done the full Saluatation to the Sun so that the whole skeleton was shown it would be more meaningful.
posted by rmmcclay at 5:24 AM on October 30, 2014 [3 favorites]
posted by rmmcclay at 5:24 AM on October 30, 2014 [3 favorites]
double block - yeah, completely. I'm on my lunch break now, but I've been doing dose calculations for medical x-rays all morning and my first thought was 'holy shit, what dose is that person getting for the sake of an advert'. After a few seconds, when I realised it was obviously CGI'd my heart rate went down a little bit.
The question of whether the dose would be fatal is an interesting one, as to get this kind of video you'd have to pull off something really odd, and whichever way you do it, you're going to push the dose up to massive levels.
At the moment, there just isn't any equipment on the market (that I'm aware of) that could do of this kind of thing, probably as there's no possible reason the benefit gained from it would outweigh the dose. You've got CT, but that doesn't cope well with movement. That's mainly because a normal CT works by taking a few slices, then moving down the patient and taking a few more, and stitching them all together. That only takes around 5 seconds, but still, it's not good enough for a moving person. The other modality that can pull off vaguely similar stuff to this is fluoroscopy - think of a (modern version of a) cartoon of an x-ray where someone walks behind a screen and you can see their bones in real time. The problem with that is you get a 2D image rather than a 3D one. The final thing that's different here to any medical exposure, is that the image quality is so good. For diagnostic purposes, there's very little point in having that level of resolution or such a high frame rate, and by doing so, you're massively pushing up the dose.
Anyway, what would the dose be? It depends on how you get the image. There are some clever systems that sort of do CT stuff with fluoroscopy sets, but they don't have anywhere near this level of performance, so I'll ignore those. If you wanted to do a 2D fluro video of this, at a normal frame rate and resolution, doses of around 0.05 Gy for the 2 minutes are quoted. However (ignoring resolution) that's at 4 frames per second, and a normal fluro set will take images up to 22cm across. By just increasing the frame rate to 20 fps, and the field size to a meter, the dose goes to 5 Gy, which is about a lethal dose*. And this is still for a 2D image with a low (spatial) resolution. If you wanted a 3D image, there are lots of ways of taking lots of 2D projections of the same thing and reconstructing a 3D image, but at that point we're basically reinventing CT, so let's look at that.
The dose from a whole body CT is quoted at around 20mGy. If we completely ignore any engineering requirements, and just state we're going to take twenty of them a second, the total dose is 48Gy. 48Gy is a really massive, a die in minutes kind of dose. One interesting thing about that value, is that if you took ten fluroscopic projections (which is the right sort of ball-park figure), you could make a 3D fluro video, and the dose would be 50Gy, pretty much the same as the 48Gy figure here. That's good, because although the starting point for both calculations is different, in both you're doing essentially the same thing: taking x-rays from a lot of angles at 20fps, and sticking them together to make a video. But anyway, even though all my estimates were a bit conservative, it's still a lethal dose, even ten seconds of that video would probably be enough to kill you. TL:DR you'd die.
*The dose at which half of people will die of radiation exposure in 30 days (the LD50/30) is between 4 and 5 Gray (Gy).
posted by Ned G at 6:10 AM on October 30, 2014 [11 favorites]
The question of whether the dose would be fatal is an interesting one, as to get this kind of video you'd have to pull off something really odd, and whichever way you do it, you're going to push the dose up to massive levels.
At the moment, there just isn't any equipment on the market (that I'm aware of) that could do of this kind of thing, probably as there's no possible reason the benefit gained from it would outweigh the dose. You've got CT, but that doesn't cope well with movement. That's mainly because a normal CT works by taking a few slices, then moving down the patient and taking a few more, and stitching them all together. That only takes around 5 seconds, but still, it's not good enough for a moving person. The other modality that can pull off vaguely similar stuff to this is fluoroscopy - think of a (modern version of a) cartoon of an x-ray where someone walks behind a screen and you can see their bones in real time. The problem with that is you get a 2D image rather than a 3D one. The final thing that's different here to any medical exposure, is that the image quality is so good. For diagnostic purposes, there's very little point in having that level of resolution or such a high frame rate, and by doing so, you're massively pushing up the dose.
Anyway, what would the dose be? It depends on how you get the image. There are some clever systems that sort of do CT stuff with fluoroscopy sets, but they don't have anywhere near this level of performance, so I'll ignore those. If you wanted to do a 2D fluro video of this, at a normal frame rate and resolution, doses of around 0.05 Gy for the 2 minutes are quoted. However (ignoring resolution) that's at 4 frames per second, and a normal fluro set will take images up to 22cm across. By just increasing the frame rate to 20 fps, and the field size to a meter, the dose goes to 5 Gy, which is about a lethal dose*. And this is still for a 2D image with a low (spatial) resolution. If you wanted a 3D image, there are lots of ways of taking lots of 2D projections of the same thing and reconstructing a 3D image, but at that point we're basically reinventing CT, so let's look at that.
The dose from a whole body CT is quoted at around 20mGy. If we completely ignore any engineering requirements, and just state we're going to take twenty of them a second, the total dose is 48Gy. 48Gy is a really massive, a die in minutes kind of dose. One interesting thing about that value, is that if you took ten fluroscopic projections (which is the right sort of ball-park figure), you could make a 3D fluro video, and the dose would be 50Gy, pretty much the same as the 48Gy figure here. That's good, because although the starting point for both calculations is different, in both you're doing essentially the same thing: taking x-rays from a lot of angles at 20fps, and sticking them together to make a video. But anyway, even though all my estimates were a bit conservative, it's still a lethal dose, even ten seconds of that video would probably be enough to kill you. TL:DR you'd die.
*The dose at which half of people will die of radiation exposure in 30 days (the LD50/30) is between 4 and 5 Gray (Gy).
posted by Ned G at 6:10 AM on October 30, 2014 [11 favorites]
And in case that post makes x-rays sound really scary and dangerous, I'd like to clarify: taking that kind of dataset is completely insane from a medical point of view, as it would involve taking 2,400 individual whole body CT scans. This is the kind of dose you'd get from having 100 chest x-rays taken every day of your life, so no need to cancel your next doctors appointment.
posted by Ned G at 6:18 AM on October 30, 2014 [3 favorites]
posted by Ned G at 6:18 AM on October 30, 2014 [3 favorites]
« Older Real Amazons wore pants | If everyone sees your dickpic hanging in a gallery... Newer »
This thread has been archived and is closed to new comments
posted by JujuB at 12:01 AM on October 30, 2014 [1 favorite]