WEBVTT Kind: captions Language: en 1 00:00:00.120 --> 00:00:05.540 In 2003, the Human Genome Project finished a monumental scientific achievement: 2 00:00:05.540 --> 00:00:09.460 the first complete human genome map. 3 00:00:09.460 --> 00:00:13.059 A couple of decades later, and you can get a genetic test done 4 00:00:13.059 --> 00:00:16.930 by spitting in a tube and paying a couple of hundred dollars. 5 00:00:16.930 --> 00:00:19.830 But that's not quite what I'm here to talk about. 6 00:00:19.830 --> 00:00:24.050 This is CUBRIC, one of the most advanced brain imaging centres in Europe, 7 00:00:24.050 --> 00:00:30.900 and here they’re not scanning genomes: they’re scanning connectomes. 8 00:00:30.900 --> 00:00:35.329 The 'connectome' is a phrase that was coined several years ago 9 00:00:35.329 --> 00:00:37.460 and it's to really encapsulate the idea of 10 00:00:37.460 --> 00:00:40.120 mapping all the connections in the brain, in one go. 11 00:00:40.120 --> 00:00:43.210 Those connections are formed by what we call the white matter in the brain. 12 00:00:43.210 --> 00:00:44.910 So we know the brain has grey matter, 13 00:00:44.910 --> 00:00:45.920 it also has white matter, 14 00:00:45.920 --> 00:00:47.239 and these form these wires. 15 00:00:47.239 --> 00:00:50.260 We're also really interested in what we call the microstructure, 16 00:00:50.260 --> 00:00:53.760 being able to differentiate between the size and shapes of cells 17 00:00:53.760 --> 00:00:56.580 that may be only a thousand of a millimetre across. 18 00:00:56.589 --> 00:00:58.249 What this scanner will enable us to do 19 00:00:58.249 --> 00:01:01.589 is to get information about differences across the whole brain 20 00:01:01.589 --> 00:01:04.600 in a way that would only previously be possible post-mortem. 21 00:01:04.600 --> 00:01:06.740 Now, I’d be getting a scan myself, but alas, 22 00:01:06.740 --> 00:01:09.680 I have a bit too much metal in my body. 23 00:01:09.680 --> 00:01:12.070 Don't worry about me or the camera getting sucked into this, though, 24 00:01:12.070 --> 00:01:13.860 this is just the simulator that they train on. 25 00:01:13.860 --> 00:01:16.050 There's no actual magnets in there. 26 00:01:16.050 --> 00:01:17.990 The real scanner is over there. 27 00:01:17.990 --> 00:01:20.590 This MRI scanner, the magnetic field gradient allows you 28 00:01:20.590 --> 00:01:23.250 to be sensitive to the motion of water molecules 29 00:01:23.250 --> 00:01:26.210 over a distance of about a thousandth of a millimetre. 30 00:01:26.210 --> 00:01:28.210 To map somebody's complete connectome, 31 00:01:28.210 --> 00:01:30.130 first we have to acquire the raw data, 32 00:01:30.130 --> 00:01:32.210 then we have to fit it to mathematical models, 33 00:01:32.210 --> 00:01:34.330 then we have to reconstruct those pathways. 34 00:01:34.330 --> 00:01:38.020 And so we rapidly expand from a few hundreds of megabytes 35 00:01:38.020 --> 00:01:39.800 up to tens of gigabytes. 36 00:01:39.800 --> 00:01:43.930 All the hopeful science fiction fans out there shouldn’t get too excited: 37 00:01:43.930 --> 00:01:47.990 in the same way as copying someone’s genome doesn't let you make an actual copy of them, 38 00:01:47.990 --> 00:01:52.150 just a baby that’s an identical twin and that doesn't know anything, 39 00:01:52.150 --> 00:01:56.750 scanning a connectome doesn’t mean that you can suddenly upload someone into a computer. 40 00:01:56.750 --> 00:02:00.730 Scanners like this generate road maps, not satellite photos. 41 00:02:00.730 --> 00:02:01.900 Heck of a road map, though. 42 00:02:01.900 --> 00:02:03.490 I think we're a long way from being able 43 00:02:03.490 --> 00:02:07.060 to identify differences in personality at the present time, 44 00:02:07.060 --> 00:02:09.119 but that's not to say we won't be able to do it in the future. 45 00:02:09.119 --> 00:02:11.129 Generally, people are connected in the same ways, 46 00:02:11.129 --> 00:02:13.680 but it's subtle differences in the shapes of the tracks 47 00:02:13.680 --> 00:02:15.069 that does seem to be highly reproducible. 48 00:02:15.069 --> 00:02:20.650 If we scan 300 people, and then scan one of those people just one more time, 49 00:02:20.650 --> 00:02:23.769 you can easily identify them out of that 300 people. 50 00:02:23.769 --> 00:02:27.810 Essentially, any neurological or psychiatric disease that you can think of, 51 00:02:27.810 --> 00:02:30.540 they're all symptoms of malfunction of the brain. 52 00:02:30.540 --> 00:02:34.049 And we know that function of the brain relies on information transfer 53 00:02:34.049 --> 00:02:35.530 from one region to another. 54 00:02:35.530 --> 00:02:40.140 We can say things such as "there's something different with the integrity of the white matter". 55 00:02:40.140 --> 00:02:43.360 Potentially, if you can name a neurological or psychiatric disease, 56 00:02:43.360 --> 00:02:45.200 the information that we get from this scanner 57 00:02:45.200 --> 00:02:47.819 is going to be a boon and give us major new insights. 58 00:02:47.819 --> 00:02:50.060 And if we could bring the cost of the technology down, 59 00:02:50.060 --> 00:02:52.739 you could go in for your connectome scan, 60 00:02:52.739 --> 00:02:57.200 and have the personalised treatment tailored to your specific connectomic deficit. 61 00:02:57.200 --> 00:03:00.290 Right now connectome scans are cumbersome and expensive, 62 00:03:00.290 --> 00:03:03.430 and they take a long time to process and visualise. 63 00:03:03.430 --> 00:03:07.189 But that was true of genome scans just 20 years ago. 64 00:03:07.189 --> 00:03:12.720 In a decade or two, connectome research might unlock just as many benefits for the world. 65 00:03:14.249 --> 00:03:17.180 Thank you very much to everyone at CUBRIC, at Cardiff University. 66 00:03:17.189 --> 00:03:20.049 If you want to know more, they have a YouTube channel you can check out, 67 00:03:20.049 --> 00:03:21.489 or there's a link in the description.