[Script Info]
Title: 
[Events]
Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
Dialogue: 0,0:00:00.00,0:00:15.64,Default,,0000,0000,0000,,{\i1}Music{\i0}
Dialogue: 0,0:00:18.16,0:00:24.94,Default,,0000,0000,0000,,Herald Angel: And now we come to the talk\Nentitled low-cost non-invasive biomedical
Dialogue: 0,0:00:24.94,0:00:32.00,Default,,0000,0000,0000,,imaging. Current medical imaging has\Nproblems: it is expensive, it is large,
Dialogue: 0,0:00:32.00,0:00:39.94,Default,,0000,0000,0000,,rarely preventively used and maybe you've\Nheard of the story of a fMRI - this is the
Dialogue: 0,0:00:39.94,0:00:45.96,Default,,0000,0000,0000,,magnet resonance tomography - they put in\Na dead Salmon and they can get a signal
Dialogue: 0,0:00:45.96,0:00:51.73,Default,,0000,0000,0000,,from brain activity from it. There's also\Nlots of problems in the software as well.
Dialogue: 0,0:00:51.73,0:00:59.45,Default,,0000,0000,0000,,A little story, maybe you look it up. And\Nhow this whole mess can be solved with the
Dialogue: 0,0:00:59.45,0:01:06.77,Default,,0000,0000,0000,,technique called Open Electrical Impedance\NTomography - this will tell us Jean
Dialogue: 0,0:01:06.77,0:01:10.51,Default,,0000,0000,0000,,Rintoul. Give a big round of applause for\NJean.
Dialogue: 0,0:01:10.51,0:01:18.54,Default,,0000,0000,0000,,{\i1}applause{\i0}\NJean Rintoul: Thank you.
Dialogue: 0,0:01:18.54,0:01:20.94,Default,,0000,0000,0000,,Hello everyone. Today I
Dialogue: 0,0:01:20.94,0:01:28.16,Default,,0000,0000,0000,,will be talking about an open source route\Nfor biomedical imaging using a technique
Dialogue: 0,0:01:28.16,0:01:34.09,Default,,0000,0000,0000,,that's in R&D called Electrical Impedance\NTomography. Not many people have heard of
Dialogue: 0,0:01:34.09,0:01:42.61,Default,,0000,0000,0000,,it, which is why it seems like it's\Nimportant to mention. First of all, I'll
Dialogue: 0,0:01:42.61,0:01:48.79,Default,,0000,0000,0000,,just give you the vision of what it would\Nbe like if everybody had access to cheap
Dialogue: 0,0:01:48.79,0:01:56.21,Default,,0000,0000,0000,,biomedical imaging. Right now you only get\Nimaged when something's gone wrong. And,
Dialogue: 0,0:01:56.21,0:02:02.19,Default,,0000,0000,0000,,moreover, you only actually get to use\Nthese tools when something has gone wrong
Dialogue: 0,0:02:02.19,0:02:08.78,Default,,0000,0000,0000,,in a first world country when you're lucky\Nenough to be close to a hospital and have
Dialogue: 0,0:02:08.78,0:02:15.03,Default,,0000,0000,0000,,access to these technologies. That's a\Nvery limited number of people. What's even
Dialogue: 0,0:02:15.03,0:02:20.72,Default,,0000,0000,0000,,worse about it: is it's hard to hack! So,\Nif you wanted to improve this technology
Dialogue: 0,0:02:20.72,0:02:28.26,Default,,0000,0000,0000,,yourself - medical physics is an amazing\Nfield - but it would be very hard to do so
Dialogue: 0,0:02:28.26,0:02:34.51,Default,,0000,0000,0000,,because you don't have a three million\Ndollar MRI scanner sitting in your garage.
Dialogue: 0,0:02:34.51,0:02:40.36,Default,,0000,0000,0000,,Maybe you do, that's good for you, just\Nnot many of us do. If we did have cheap
Dialogue: 0,0:02:40.36,0:02:46.19,Default,,0000,0000,0000,,biomedical imaging we could do things like\Ndo preventive scans so you would wake up
Dialogue: 0,0:02:46.19,0:02:51.52,Default,,0000,0000,0000,,in the morning you'd like, take a shower,\Nthe device would be quietly imaging your
Dialogue: 0,0:02:51.52,0:02:55.75,Default,,0000,0000,0000,,body, would warn you if the slightest\Nlittle thing when went wrong. You'd do
Dialogue: 0,0:02:55.75,0:03:02.29,Default,,0000,0000,0000,,machine learning over it, it'd be\Nwonderful wonderful for health care. So,
Dialogue: 0,0:03:02.29,0:03:07.75,Default,,0000,0000,0000,,that's the vision of what biomedical\Nimaging could be. And the other point is
Dialogue: 0,0:03:07.75,0:03:13.31,Default,,0000,0000,0000,,sometimes we move forward faster when we\Nshare the information. I worked in defense
Dialogue: 0,0:03:13.31,0:03:16.41,Default,,0000,0000,0000,,for a brief period and people didn't\Nreally share information between each
Dialogue: 0,0:03:16.41,0:03:21.44,Default,,0000,0000,0000,,other, and I think that inhibited science\Nfrom moving forward. So, sharing is
Dialogue: 0,0:03:21.44,0:03:24.76,Default,,0000,0000,0000,,caring.\NSo today I'm going to go through a few
Dialogue: 0,0:03:24.76,0:03:28.02,Default,,0000,0000,0000,,different things. I'm going to go through\Nthe current biomedical imaging
Dialogue: 0,0:03:28.02,0:03:31.02,Default,,0000,0000,0000,,technologies. I'll give you an\Nintroduction to Electrical Impedance
Dialogue: 0,0:03:31.02,0:03:35.59,Default,,0000,0000,0000,,Tomography. I'll go through the open\Nsource Electrical Impedance Tomography
Dialogue: 0,0:03:35.59,0:03:40.66,Default,,0000,0000,0000,,Project. Then I'll go through some\Napplications that we could apply it to.
Dialogue: 0,0:03:40.66,0:03:45.04,Default,,0000,0000,0000,,And then I'll suggest a few different next\Nsteps that we can go into because by no
Dialogue: 0,0:03:45.04,0:03:52.98,Default,,0000,0000,0000,,means is it finished. Right now we have\Nfour different main existing imaging
Dialogue: 0,0:03:52.98,0:04:00.91,Default,,0000,0000,0000,,modalities. Your MRI scanner, which is a\Nwonderful tool, it's huge, very expensive.
Dialogue: 0,0:04:00.91,0:04:06.32,Default,,0000,0000,0000,,The most commonly used imaging is actually\NCAT scanner which sends our x-rays through
Dialogue: 0,0:04:06.32,0:04:11.07,Default,,0000,0000,0000,,your body which is ionizing radiation,\Nwhich is bad for you because it causes
Dialogue: 0,0:04:11.07,0:04:17.22,Default,,0000,0000,0000,,cancer in the long run if you get too many\Nof those scans and it's actually the first
Dialogue: 0,0:04:17.22,0:04:21.46,Default,,0000,0000,0000,,first scan that you'll get when you go\Ninto the emergency room. It's the most
Dialogue: 0,0:04:21.46,0:04:24.65,Default,,0000,0000,0000,,commonly used. And as we all know we've\Ngot those grainy images that come from the
Dialogue: 0,0:04:24.65,0:04:32.40,Default,,0000,0000,0000,,ultrasound of fetuses, wonderful tool\Nexcept for the scattering due to the sound
Dialogue: 0,0:04:32.40,0:04:36.30,Default,,0000,0000,0000,,gets scattered when you have different\Ndensity materials next to each other. And
Dialogue: 0,0:04:36.30,0:04:45.16,Default,,0000,0000,0000,,not exactly an imaging modality but a very\Nimportant diagnostic technique is EEG.
Dialogue: 0,0:04:45.16,0:04:50.98,Default,,0000,0000,0000,,So you might ask, how do we classify these\Nright now? we have 3 main types of
Dialogue: 0,0:04:50.98,0:04:58.85,Default,,0000,0000,0000,,resolution. Spatial, contrast, and time.\NSpatial resolution is, basically, what
Dialogue: 0,0:04:58.85,0:05:04.01,Default,,0000,0000,0000,,space you can determine 2 different\Nobjects from each other. Contrast
Dialogue: 0,0:05:04.01,0:05:09.00,Default,,0000,0000,0000,,resolution is soft tissue or subtle\Ndifferences in tissues. And time
Dialogue: 0,0:05:09.00,0:05:13.54,Default,,0000,0000,0000,,resolution, as it sounds, is how things\Nchange over time and how quickly you can
Dialogue: 0,0:05:13.54,0:05:19.40,Default,,0000,0000,0000,,do these images together. Your CAT scan,\Nyour basic machine in a hospital,
Dialogue: 0,0:05:19.40,0:05:24.21,Default,,0000,0000,0000,,costs 1 to 2.5 million dollars.\NYou probably didn't get one for Christmas
Dialogue: 0,0:05:24.21,0:05:30.02,Default,,0000,0000,0000,,to play around with. Oh well. It's also\Ngot this ionizing radiation, you've got
Dialogue: 0,0:05:30.02,0:05:32.71,Default,,0000,0000,0000,,a lot of maintenance, and \Ndedicated technicians.
Dialogue: 0,0:05:32.71,0:05:35.24,Default,,0000,0000,0000,,An MRI, say, your average 3 Tesla magnet
Dialogue: 0,0:05:35.24,0:05:41.06,Default,,0000,0000,0000,,with its own helium quenching chamber\Nno less, as well as dedicated technicians
Dialogue: 0,0:05:41.06,0:05:50.71,Default,,0000,0000,0000,,and experts who can actually read\Nthe images. Again $3,000,000. An amazing
Dialogue: 0,0:05:50.71,0:05:55.95,Default,,0000,0000,0000,,and beautiful technology, but really\Nexpensive. Amazing spatial resolution, the
Dialogue: 0,0:05:55.95,0:06:01.01,Default,,0000,0000,0000,,best. When it does something at this very\Nhigh spatial resolution, it actually takes
Dialogue: 0,0:06:01.01,0:06:06.49,Default,,0000,0000,0000,,4 minutes and 16 seconds. Which is a\Nreally long time to take to do this
Dialogue: 0,0:06:06.49,0:06:11.20,Default,,0000,0000,0000,,wonderful spatial resolution image.\NUltrasound, it's a bit grainy due to
Dialogue: 0,0:06:11.20,0:06:18.56,Default,,0000,0000,0000,,scattering. On average it costs about\N1$115k, not too bad. It's a pretty minimal
Dialogue: 0,0:06:18.56,0:06:25.46,Default,,0000,0000,0000,,health risk. EEG. EEG doesn't do any image\Nreconstruction. In fact it does very
Dialogue: 0,0:06:25.46,0:06:33.04,Default,,0000,0000,0000,,little in many ways. But it is still very\Nuseful. Your average medical grade by EEG
Dialogue: 0,0:06:33.04,0:06:37.92,Default,,0000,0000,0000,,system is $40k. You might also know of\Nsome open source EEG projects which are
Dialogue: 0,0:06:37.92,0:06:45.07,Default,,0000,0000,0000,,pretty cool. So just a note on the\Nradiation of CAT-scans. It's actually the
Dialogue: 0,0:06:45.07,0:06:53.70,Default,,0000,0000,0000,,biggest contributing cause of radiation in\Nthe United States. So here I just put
Dialogue: 0,0:06:53.70,0:06:58.55,Default,,0000,0000,0000,,those biomedical imaging modalities onto a\Ngraph so that you can kind of think of
Dialogue: 0,0:06:58.55,0:07:03.12,Default,,0000,0000,0000,,them in terms of spatial resolution and\Ntime resolution, and where they fall in
Dialogue: 0,0:07:03.12,0:07:09.24,Default,,0000,0000,0000,,the picture of common things that go wrong\Nwith people. Like, X-rays or CAT scans are
Dialogue: 0,0:07:09.24,0:07:15.43,Default,,0000,0000,0000,,great for for looking at bone and bone\Nbreaks; pulmonary edema, that's water on
Dialogue: 0,0:07:15.43,0:07:20.76,Default,,0000,0000,0000,,the lung ,tuberculosis, huge in third-\Nworld countries, massive problem. You
Dialogue: 0,0:07:20.76,0:07:25.38,Default,,0000,0000,0000,,don't actually need super high spatial\Nresolution to be able to detect it. And
Dialogue: 0,0:07:25.38,0:07:29.61,Default,,0000,0000,0000,,it's important to sort of understand what\Nyou can do at different spatial and time
Dialogue: 0,0:07:29.61,0:07:35.43,Default,,0000,0000,0000,,resolutions. Under like, the optimal goal\Nof all of this, I put non-invasive
Dialogue: 0,0:07:35.43,0:07:40.37,Default,,0000,0000,0000,,electrophysiology. What that is, is high\Nspatial resolution and high time
Dialogue: 0,0:07:40.37,0:07:46.82,Default,,0000,0000,0000,,resolution. That's where you can measure\Nion activation, or basically what cells
Dialogue: 0,0:07:46.82,0:07:51.46,Default,,0000,0000,0000,,are doing when they communicate with each\Nother, which is right now only done in an
Dialogue: 0,0:07:51.46,0:07:55.85,Default,,0000,0000,0000,,invasive manner.\NToday I'm gonna talk about this new
Dialogue: 0,0:07:55.85,0:08:00.87,Default,,0000,0000,0000,,technique called Electrical Impedance\NTomography and describe where it will fit
Dialogue: 0,0:08:00.87,0:08:09.65,Default,,0000,0000,0000,,in amongst what already exists. So what is\Nit. Okay yeah basically you send AC
Dialogue: 0,0:08:09.65,0:08:17.13,Default,,0000,0000,0000,,currents through the body, say a 50\Nkilohertz current. And that will take
Dialogue: 0,0:08:17.13,0:08:23.36,Default,,0000,0000,0000,,different routes based on what tissue\Nthere is. So it might go around some cells
Dialogue: 0,0:08:23.36,0:08:29.62,Default,,0000,0000,0000,,and straight through others. And that's\Nreally important because differentiating,
Dialogue: 0,0:08:29.62,0:08:34.87,Default,,0000,0000,0000,,say, fat from muscle is one thing that you\Ncould do. But you can go further and
Dialogue: 0,0:08:34.87,0:08:41.97,Default,,0000,0000,0000,,differentiate, say, tumors from healthy\Ntissue. Because tumors have different
Dialogue: 0,0:08:41.97,0:08:47.96,Default,,0000,0000,0000,,impedance spectra to the healthy tissue.\NSo as you can see, that would be very
Dialogue: 0,0:08:47.96,0:08:53.47,Default,,0000,0000,0000,,useful to do. This set up here is a called\Na phantom. What it is, it's like a
Dialogue: 0,0:08:53.47,0:08:58.65,Default,,0000,0000,0000,,simulated human body. You get some\Nsaltwater - the body is 80% water as you
Dialogue: 0,0:08:58.65,0:09:03.37,Default,,0000,0000,0000,,might know -you get some meat or\Nvegetables. You put it inside and then you
Dialogue: 0,0:09:03.37,0:09:07.25,Default,,0000,0000,0000,,use that to image. So we have current\Nflowing through all these different
Dialogue: 0,0:09:07.25,0:09:13.12,Default,,0000,0000,0000,,directions and we recreate an image. Right\Nnow it's used for lung volume
Dialogue: 0,0:09:13.12,0:09:18.40,Default,,0000,0000,0000,,measurements. This is a baby with an EIT\Nsetup. Muscle and fat mass, there's a
Dialogue: 0,0:09:18.40,0:09:22.18,Default,,0000,0000,0000,,paper on gestural recognition that just\Ncame out this year, you can look at
Dialogue: 0,0:09:22.18,0:09:27.26,Default,,0000,0000,0000,,bladder and stomach fullness. There's some\Nresearch papers on breast and kidney
Dialogue: 0,0:09:27.26,0:09:33.86,Default,,0000,0000,0000,,cancer detection. There's another research\Npaper on hemorrhage detection for stroke.
Dialogue: 0,0:09:33.86,0:09:39.11,Default,,0000,0000,0000,,You can also look at the ... there's more\NR&D on the depth of anesthesia in in
Dialogue: 0,0:09:39.11,0:09:42.82,Default,,0000,0000,0000,,surgery as well, which would be another\Ninteresting use for it. So all of these
Dialogue: 0,0:09:42.82,0:09:50.65,Default,,0000,0000,0000,,are sort of in the works and you might\Nask, "Great, that sounds amazing, why
Dialogue: 0,0:09:50.65,0:09:55.59,Default,,0000,0000,0000,,isn't everybody using it already?" Well\Nyeah it's really an R&D technique right
Dialogue: 0,0:09:55.59,0:10:03.03,Default,,0000,0000,0000,,now and it has a big problem: its spatial\Nresolution seems pretty limited. So it's
Dialogue: 0,0:10:03.03,0:10:06.99,Default,,0000,0000,0000,,limited by the number of electrodes. But I\Nwill discuss some potential ways to get
Dialogue: 0,0:10:06.99,0:10:12.93,Default,,0000,0000,0000,,around that. As we go, it might not ever\Nget to the spatial resolution of MRI.
Dialogue: 0,0:10:12.93,0:10:16.74,Default,,0000,0000,0000,,But maybe we don't need it to to be\Nuseful. Because it's so compact. It's so
Dialogue: 0,0:10:16.74,0:10:23.81,Default,,0000,0000,0000,,cheap, nothing about it is expensive. It's\Ngot better source localization than EEG.
Dialogue: 0,0:10:23.81,0:10:27.76,Default,,0000,0000,0000,,It does not ionize,\Nit's not harmful to human tissue. It's
Dialogue: 0,0:10:27.76,0:10:33.59,Default,,0000,0000,0000,,also got great time resolution, so it has\Nadvantages and disadvantages. I'll just
Dialogue: 0,0:10:33.59,0:10:39.24,Default,,0000,0000,0000,,remind you of what the first MRI scan\Nlooked like at this point in time. As you
Dialogue: 0,0:10:39.24,0:10:45.60,Default,,0000,0000,0000,,can see it looks pretty crappy in 1977.\NAnd now it looks pretty awesome. That's a
Dialogue: 0,0:10:45.60,0:10:51.23,Default,,0000,0000,0000,,slice of my head by the way in a 3 Tesla\NMRI scanner. This is what early EIT looks
Dialogue: 0,0:10:51.23,0:10:59.43,Default,,0000,0000,0000,,like. That's with 16 electrodes only. What\Nwill it look like in a few years time I
Dialogue: 0,0:10:59.43,0:11:06.90,Default,,0000,0000,0000,,don't know. I hope that MRI gives you a\Npathway that it will take take too.
Dialogue: 0,0:11:06.90,0:11:13.11,Default,,0000,0000,0000,,Now I'll introduce you to the OpenEIT\Nproject. The OpenEIT project is obviously
Dialogue: 0,0:11:13.11,0:11:20.45,Default,,0000,0000,0000,,open source. It has a PCB design done in\NEagle CAD. It has firmware written in C.
Dialogue: 0,0:11:20.45,0:11:24.92,Default,,0000,0000,0000,,It has a Python dashboard that lets you\Nsee the reconstruction in real time. It
Dialogue: 0,0:11:24.92,0:11:29.30,Default,,0000,0000,0000,,also has a reconstruction algorithm which\NI'll go into. And you can get it from
Dialogue: 0,0:11:29.30,0:11:36.54,Default,,0000,0000,0000,,github right there. So how does it\Nreconstruct an image? OpenEIT right now
Dialogue: 0,0:11:36.54,0:11:43.06,Default,,0000,0000,0000,,has 8 electrodes and what you do is, you\Nsend this 50 kHz current through every
Dialogue: 0,0:11:43.06,0:11:48.92,Default,,0000,0000,0000,,combination of those 8 electrodes and you\Nget a different impedance value for each
Dialogue: 0,0:11:48.92,0:11:55.52,Default,,0000,0000,0000,,of those measurements. On the left you can\Nsee basically what you're doing. You know where the
Dialogue: 0,0:11:55.52,0:12:01.27,Default,,0000,0000,0000,,electrodes are positioned and you get one\Nvalue going horizontally. You add it to
Dialogue: 0,0:12:01.27,0:12:06.10,Default,,0000,0000,0000,,another value coming from another\Ndirection. And again, you can sort of see
Dialogue: 0,0:12:06.10,0:12:11.24,Default,,0000,0000,0000,,it's getting a low resolution image as it\Ngoes around adding those values together.
Dialogue: 0,0:12:11.24,0:12:18.37,Default,,0000,0000,0000,,If you use many, many views you bring the\Nimage back. This is the radon transform,
Dialogue: 0,0:12:18.37,0:12:23.75,Default,,0000,0000,0000,,that's what it's called, and you \Nbasically just send lots of current
Dialogue: 0,0:12:23.75,0:12:26.90,Default,,0000,0000,0000,,through these different slightly different\Nangles and you build up something called a
Dialogue: 0,0:12:26.90,0:12:33.58,Default,,0000,0000,0000,,sinogram which is over there. And then you\Ninvert it to get the image back. I used
Dialogue: 0,0:12:33.58,0:12:37.21,Default,,0000,0000,0000,,OpenCV which is a really common image\Nprocessing library to do this. You can
Dialogue: 0,0:12:37.21,0:12:43.36,Default,,0000,0000,0000,,just do it with a regular image yourself\Nand try it out. But what I did is exactly
Dialogue: 0,0:12:43.36,0:12:49.07,Default,,0000,0000,0000,,the same as what you do with a regular\Nimage, except I use current to be the
Dialogue: 0,0:12:49.07,0:12:57.81,Default,,0000,0000,0000,,input data. So this is the PCB design \Nin Eagle. Basically it has a
Dialogue: 0,0:12:57.81,0:13:02.83,Default,,0000,0000,0000,,few different features. A connector for\Nyour 8 electrodes. It's running an ARM
Dialogue: 0,0:13:02.83,0:13:11.34,Default,,0000,0000,0000,,Cortex M3, which is quite nice. It has a\Ndedicated DFT engine for doing your direct
Dialogue: 0,0:13:11.34,0:13:16.83,Default,,0000,0000,0000,,Fourier transform in real, time which is\Nalso quite nice. A JTAG debugger to easily
Dialogue: 0,0:13:16.83,0:13:22.62,Default,,0000,0000,0000,,reprogram it. It's got coin cell or\Nexternal battery options. It has UART to
Dialogue: 0,0:13:22.62,0:13:28.66,Default,,0000,0000,0000,,get the serial data off. And you can also\Nflip it to Bluetooth mode and get the data
Dialogue: 0,0:13:28.66,0:13:32.04,Default,,0000,0000,0000,,off by Bluetooth if you felt like going\NWireless.
Dialogue: 0,0:13:32.04,0:13:37.13,Default,,0000,0000,0000,,At this point you might be asking "Is this\Nsafe for me to play around with?", which
Dialogue: 0,0:13:37.13,0:13:42.88,Default,,0000,0000,0000,,is a really great question because the\Nanswer is actually "Yeah! it is". There's
Dialogue: 0,0:13:42.88,0:13:51.05,Default,,0000,0000,0000,,some guidelines called the IEC60601-1\Nguidelines for safer use in humans. And
Dialogue: 0,0:13:51.05,0:13:56.95,Default,,0000,0000,0000,,basically which says it should be, and\NopenEIT is less than 10 micro amps which
Dialogue: 0,0:13:56.95,0:14:02.96,Default,,0000,0000,0000,,is great because that's well within their\Nguidelines. If you want to compare it to
Dialogue: 0,0:14:02.96,0:14:06.22,Default,,0000,0000,0000,,other things that are completely legal,\Nsay I don't know if you've seen there's
Dialogue: 0,0:14:06.22,0:14:10.63,Default,,0000,0000,0000,,like late-night TV ads for those abs\Nstimulators that stimulate your muscles,
Dialogue: 0,0:14:10.63,0:14:17.27,Default,,0000,0000,0000,,there are about 15 to 20 milliamps just\Nfor reference and as a scale to look at
Dialogue: 0,0:14:17.27,0:14:23.75,Default,,0000,0000,0000,,the 10 micro amps. So some of you might\Nhave used them already and that's hugely
Dialogue: 0,0:14:23.75,0:14:27.89,Default,,0000,0000,0000,,more current than what we're putting\Nthrough to image the body here. This is
Dialogue: 0,0:14:27.89,0:14:33.48,Default,,0000,0000,0000,,what the dashboard looks like. It does the\Nreconstruction. You can connect to serial
Dialogue: 0,0:14:33.48,0:14:37.85,Default,,0000,0000,0000,,at baseline. You can obviously adjust\Nsliders to look at the area that you want
Dialogue: 0,0:14:37.85,0:14:43.85,Default,,0000,0000,0000,,to look at. You can read from a file and\Nfiddle around however you would like to.
Dialogue: 0,0:14:43.85,0:14:49.17,Default,,0000,0000,0000,,This is what it looks like when you\Nreconstruct something. I have a phantom up
Dialogue: 0,0:14:49.17,0:14:54.08,Default,,0000,0000,0000,,there which is a part of water with a cup\Nin it. I moved the cup around anti-
Dialogue: 0,0:14:54.08,0:14:58.90,Default,,0000,0000,0000,,clockwise so you can see in each of the\Npictures I move it around a little bit
Dialogue: 0,0:14:58.90,0:15:04.40,Default,,0000,0000,0000,,more. And you can see the reconstruction\Nthere with me moving the cup around again.
Dialogue: 0,0:15:04.40,0:15:07.97,Default,,0000,0000,0000,,This might not be wow-ing you with the\Nresolution, with only 8 electrodes. It's a
Dialogue: 0,0:15:07.97,0:15:14.80,Default,,0000,0000,0000,,proof of concept but that's okay. Let's\Nsee if we can make this I make this go.
Dialogue: 0,0:15:14.80,0:15:20.72,Default,,0000,0000,0000,,Here's a real-time video demonstration of\Nit. Here's me with a shot glass. I'm
Dialogue: 0,0:15:20.72,0:15:24.64,Default,,0000,0000,0000,,moving around anti-clockwise. Hopefully\Nyou can see on the left the image being
Dialogue: 0,0:15:24.64,0:15:33.92,Default,,0000,0000,0000,,reconstructed in real time. And there we\Ngo, move to the bottom. You can see it
Dialogue: 0,0:15:33.92,0:15:41.03,Default,,0000,0000,0000,,over there and again up to the top. you\Ncan see it over there. So that's a basic
Dialogue: 0,0:15:41.03,0:15:44.74,Default,,0000,0000,0000,,proof of principle version of it running.
Dialogue: 0,0:15:49.13,0:15:55.08,Default,,0000,0000,0000,,So the first MRI scan of human \Nlungs wasn't that amazing.
Dialogue: 0,0:15:55.08,0:15:57.63,Default,,0000,0000,0000,,Early EIT scan wasn't either.
Dialogue: 0,0:15:57.63,0:16:02.82,Default,,0000,0000,0000,,{\i1}applause{\i0}\NSomething else that you can use it
Dialogue: 0,0:16:02.82,0:16:08.59,Default,,0000,0000,0000,,that for is differentiating objects.\NMulti-frequency. This is what they're
Dialogue: 0,0:16:08.59,0:16:13.97,Default,,0000,0000,0000,,doing the breast cancer and kidney cancer\Nscans on. Basically you send different
Dialogue: 0,0:16:13.97,0:16:17.92,Default,,0000,0000,0000,,frequencies through these times, called\Nmulti-frequency Electrical Impedance
Dialogue: 0,0:16:17.92,0:16:23.18,Default,,0000,0000,0000,,Tomography and you build up a spectrum.\NHere I've got an apple, a pear oh no a
Dialogue: 0,0:16:23.18,0:16:27.66,Default,,0000,0000,0000,,sweet potato and and some water. And I've\Nsent through these different frequencies
Dialogue: 0,0:16:27.66,0:16:32.31,Default,,0000,0000,0000,,and I get these different spectrums.\NThey're different, you can see that
Dialogue: 0,0:16:32.31,0:16:35.83,Default,,0000,0000,0000,,they're different. They're quite obviously\Ndifferent but yeah you can also just
Dialogue: 0,0:16:35.83,0:16:39.68,Default,,0000,0000,0000,,simply classify. And on the left you can\Nsee where the water is, the apple is, the
Dialogue: 0,0:16:39.68,0:16:44.77,Default,,0000,0000,0000,,sweet potato is. Or, the sweet potato and\Nthe apple a little bit harder that one.
Dialogue: 0,0:16:44.77,0:16:53.80,Default,,0000,0000,0000,,But that's basically what you do when you\Ndetect cancer. So that's what I did. But
Dialogue: 0,0:16:53.80,0:16:57.76,Default,,0000,0000,0000,,maybe we should look at the other papers\Nand see what they did because they did
Dialogue: 0,0:16:57.76,0:17:04.59,Default,,0000,0000,0000,,better than me. So there's this guy called\NAristovich, 2014 he published spatial and
Dialogue: 0,0:17:04.59,0:17:07.57,Default,,0000,0000,0000,,temporal resolution, and using this\Ntechnique 200 micro meters less than 2
Dialogue: 0,0:17:07.57,0:17:14.23,Default,,0000,0000,0000,,milliseconds which covers most of the\Napplications that I listed on that graph
Dialogue: 0,0:17:14.23,0:17:19.26,Default,,0000,0000,0000,,at the start of the talk. The downside\Nhere is that it was an intracranial array,
Dialogue: 0,0:17:19.26,0:17:24.19,Default,,0000,0000,0000,,so it was under the skull. So very dense\Nelectrodes, a lot more electrodes. I only
Dialogue: 0,0:17:24.19,0:17:32.09,Default,,0000,0000,0000,,used 8 he used like 256 so you can see\Nthat it can be, like, the potential is
Dialogue: 0,0:17:32.09,0:17:36.86,Default,,0000,0000,0000,,there.\NSo how should we use it first? what's a
Dialogue: 0,0:17:36.86,0:17:41.23,Default,,0000,0000,0000,,nice low hanging through fruit? What about\Nmedical imaging in the developing world
Dialogue: 0,0:17:41.23,0:17:46.65,Default,,0000,0000,0000,,where I believe 4 billion people don't\Nhave access to medical imaging. No MRI, no
Dialogue: 0,0:17:46.65,0:17:51.52,Default,,0000,0000,0000,,CAT scans. Why is the EIT good for that?\NIt's cheap to mass-produce, super
Dialogue: 0,0:17:51.52,0:17:58.09,Default,,0000,0000,0000,,portable, super low power. So that would\Nbe a great place to start. What could we
Dialogue: 0,0:17:58.09,0:18:05.33,Default,,0000,0000,0000,,do first? I'm going to go back to this\Nimage again and have a look. Tuberculosis
Dialogue: 0,0:18:05.33,0:18:09.11,Default,,0000,0000,0000,,affects a lot of people in the developing\Nworld and you don't need amazing spatial
Dialogue: 0,0:18:09.11,0:18:14.90,Default,,0000,0000,0000,,resolution to detect it. That would be a\Ngood one. Or what about a pulmonary edema?
Dialogue: 0,0:18:14.90,0:18:21.66,Default,,0000,0000,0000,,Pulmonary edema is water on the lung. It's\Nactually already used for that. You can
Dialogue: 0,0:18:21.66,0:18:26.64,Default,,0000,0000,0000,,quite easily see the different volume\Npresent, or the different conductivity
Dialogue: 0,0:18:26.64,0:18:33.59,Default,,0000,0000,0000,,maps it's called, of a working lung and a\Nnot so working lung right there.
Dialogue: 0,0:18:33.59,0:18:41.41,Default,,0000,0000,0000,,Next steps. So what should we do to make\Nthis technique better? What should we do
Dialogue: 0,0:18:41.41,0:18:48.20,Default,,0000,0000,0000,,for OpenEIT to make it better? If you want\Nto innovate again, that's the github
Dialogue: 0,0:18:48.20,0:18:53.46,Default,,0000,0000,0000,,project. Just go ahead. Oh that's an\Navocado, it has a seat in the middle. Who
Dialogue: 0,0:18:53.46,0:19:05.52,Default,,0000,0000,0000,,knew? I do. So I see the two main routes forward\Nas: One would be this low-cost biomedical
Dialogue: 0,0:19:05.52,0:19:10.69,Default,,0000,0000,0000,,imaging for the developing world. You\Ncould just stick with the static imaging
Dialogue: 0,0:19:10.69,0:19:16.27,Default,,0000,0000,0000,,reconstruction because why not. you'd need\Na few more electrodes than it currently
Dialogue: 0,0:19:16.27,0:19:22.24,Default,,0000,0000,0000,,has. One of the main problems with the\Ntechnique is how you stick it to the skin.
Dialogue: 0,0:19:22.24,0:19:25.73,Default,,0000,0000,0000,,So my suggestion for that is why don't you\Njust use a water bath and stick the body
Dialogue: 0,0:19:25.73,0:19:31.93,Default,,0000,0000,0000,,part of interest in a body of water,\Nbecause water gets rid of a lot of the,
Dialogue: 0,0:19:31.93,0:19:38.39,Default,,0000,0000,0000,,it's called the contact impedance problem.\NOr, on the kind of exciting science front,
Dialogue: 0,0:19:38.39,0:19:47.23,Default,,0000,0000,0000,,you've got the advancing neuroscience\Noption. Which would be measuring both high
Dialogue: 0,0:19:47.23,0:19:50.34,Default,,0000,0000,0000,,spatial resolution and high time\Nresolution. So that's the non-invasive
Dialogue: 0,0:19:50.34,0:19:57.71,Default,,0000,0000,0000,,electrophysiology solution. Or, and that\Nwould be super awesome, there's a couple
Dialogue: 0,0:19:57.71,0:20:03.63,Default,,0000,0000,0000,,of ways forward to do that and I'm going\Nto sort of discuss each of those.
Dialogue: 0,0:20:03.63,0:20:10.17,Default,,0000,0000,0000,,So roughly there's physical configuration\Nimprovements that could be done. There's
Dialogue: 0,0:20:10.17,0:20:14.48,Default,,0000,0000,0000,,things that you can do to improve the\Nspatial resolution. There's things you can
Dialogue: 0,0:20:14.48,0:20:19.48,Default,,0000,0000,0000,,do to improve the time resolution. And\Nthis is interesting tack on at the end
Dialogue: 0,0:20:19.48,0:20:25.21,Default,,0000,0000,0000,,that I thought I'd mentioned, which is\N'write' functionality. So we're using very
Dialogue: 0,0:20:25.21,0:20:33.09,Default,,0000,0000,0000,,small currents to read an image. What if\Nwe pumped the current up a little before
Dialogue: 0,0:20:33.09,0:20:38.94,Default,,0000,0000,0000,,you know it you're writing. I think not\Ninvasive deep brain stimulation in a
Dialogue: 0,0:20:38.94,0:20:48.67,Default,,0000,0000,0000,,focused way, that would be very very cool.\NSo, contact impedance. Major problem right
Dialogue: 0,0:20:48.67,0:20:54.06,Default,,0000,0000,0000,,now, there is a well-known solution I\Nhaven't done it yet you do this thing
Dialogue: 0,0:20:54.06,0:21:00.91,Default,,0000,0000,0000,,called differential referencing, common\Nmode rejection should be done I haven't
Dialogue: 0,0:21:00.91,0:21:05.07,Default,,0000,0000,0000,,done it that's the next step. That means\Nthat it will work when you just attach it
Dialogue: 0,0:21:05.07,0:21:10.63,Default,,0000,0000,0000,,with electrodes on the body. What happens\Nis, electrodes have a like some
Dialogue: 0,0:21:10.63,0:21:16.62,Default,,0000,0000,0000,,capacitance and different amounts which\Nkind of interfere with the the measurement
Dialogue: 0,0:21:16.62,0:21:19.69,Default,,0000,0000,0000,,that you want to make which you want to be\Nvery accurate and just of your body. You
Dialogue: 0,0:21:19.69,0:21:24.53,Default,,0000,0000,0000,,don't want to include the electrode\Ninformation in there that's changing.
Dialogue: 0,0:21:24.53,0:21:30.11,Default,,0000,0000,0000,,There's a way to remove that that's well\Nknown already. Another physical
Dialogue: 0,0:21:30.11,0:21:34.87,Default,,0000,0000,0000,,configuration improvements: just increase\Nthe number of electrodes. Wonderful, now
Dialogue: 0,0:21:34.87,0:21:41.64,Default,,0000,0000,0000,,you've just improved the resolution. Or\Nthe placing the part in water. Another set
Dialogue: 0,0:21:41.64,0:21:46.70,Default,,0000,0000,0000,,of next steps would be on the mathematical\Nside. I mentioned that I use linear back
Dialogue: 0,0:21:46.70,0:21:55.76,Default,,0000,0000,0000,,projection which is a wonderful technique,\Nthat's how they do CAT scans. With X-rays
Dialogue: 0,0:21:55.76,0:21:59.31,Default,,0000,0000,0000,,that's exactly what they do.\NHowever, it makes some appalling
Dialogue: 0,0:21:59.31,0:22:06.65,Default,,0000,0000,0000,,assumptions, like parent moves and\Nstraight lines. That is not true. What you
Dialogue: 0,0:22:06.65,0:22:11.21,Default,,0000,0000,0000,,should do is get a finite element model\Nand solve Maxwell's equations because
Dialogue: 0,0:22:11.21,0:22:18.63,Default,,0000,0000,0000,,current bends around objects. Actually it\Nworks in three dimensions too which might
Dialogue: 0,0:22:18.63,0:22:23.64,Default,,0000,0000,0000,,not be all that surprising but it needs to\Nbe solved for those three dimensions which
Dialogue: 0,0:22:23.64,0:22:26.81,Default,,0000,0000,0000,,is why you just need to solve\NMaxwell's equations and
Dialogue: 0,0:22:26.81,0:22:30.90,Default,,0000,0000,0000,,create a finite element model.\NAnd there's a quite a bit of work on
Dialogue: 0,0:22:30.90,0:22:34.61,Default,,0000,0000,0000,,mathematical solutions that get higher\Nresolution.
Dialogue: 0,0:22:34.61,0:22:42.19,Default,,0000,0000,0000,,That's another improvement area. And now\Nas I mentioned this awesome new technique.
Dialogue: 0,0:22:42.19,0:22:44.65,Default,,0000,0000,0000,,Which, actualy there's a paper on\Nthis year called
Dialogue: 0,0:22:44.65,0:22:50.99,Default,,0000,0000,0000,,magneto-acoustic electical tomography.\NYou might remember
Dialogue: 0,0:22:50.99,0:22:55.58,Default,,0000,0000,0000,,the FBI rule from high school.\NWhen you have a current flowing,
Dialogue: 0,0:22:55.58,0:23:02.43,Default,,0000,0000,0000,,perpendicular to that there will be a\Nforce. Now that force, say it's vibrating
Dialogue: 0,0:23:02.43,0:23:07.31,Default,,0000,0000,0000,,with 50 kilohertz. that's the AC signal\Nthat you're sending through. Now you have
Dialogue: 0,0:23:07.31,0:23:11.46,Default,,0000,0000,0000,,a vibrating compression wave. That's\Nsound. You can pick that up with a little
Dialogue: 0,0:23:11.46,0:23:19.58,Default,,0000,0000,0000,,piezoelectric element. And that's actually\Na focus of work. From that you can get
Dialogue: 0,0:23:19.58,0:23:27.07,Default,,0000,0000,0000,,really good edge information, because as I\Nmentioned earlier, sound scatters at
Dialogue: 0,0:23:27.07,0:23:31.46,Default,,0000,0000,0000,,edges. So you would also get the\Nelectrical impedance tomography
Dialogue: 0,0:23:31.46,0:23:38.55,Default,,0000,0000,0000,,information for the tissue sensitivity.\NWhy not combine those results together and
Dialogue: 0,0:23:38.55,0:23:43.26,Default,,0000,0000,0000,,you would have a better tool. It currently\Ngets lesser resolution in the middle
Dialogue: 0,0:23:43.26,0:23:50.18,Default,,0000,0000,0000,,simply from how you every combination of\Nelectrodes just ends up having a less
Dialogue: 0,0:23:50.18,0:23:56.80,Default,,0000,0000,0000,,dense number in the middle. You can also\Ndo something as simple as increasing the
Dialogue: 0,0:23:56.80,0:24:02.05,Default,,0000,0000,0000,,power that you send through if you're game\Nto do that. This is a kind of gory
Dialogue: 0,0:24:02.05,0:24:07.96,Default,,0000,0000,0000,,picture. Right now epileptics, if they're\Nreally troubled by their problem, which
Dialogue: 0,0:24:07.96,0:24:13.46,Default,,0000,0000,0000,,they are often, they go into a hospital\Nhave their brains opened up and they
Dialogue: 0,0:24:13.46,0:24:18.58,Default,,0000,0000,0000,,stick this array on their head through\Ntheir skull. And they leave it open
Dialogue: 0,0:24:18.58,0:24:24.34,Default,,0000,0000,0000,,for a week. And they try to induce\Nseizures through sleep deprivation.
Dialogue: 0,0:24:24.34,0:24:30.44,Default,,0000,0000,0000,,And then they measure the activation\Npotentials that way to locate the foci or
Dialogue: 0,0:24:30.44,0:24:36.20,Default,,0000,0000,0000,,where they going to do surgery to stop you\Nfrom having seizures. But it would be much
Dialogue: 0,0:24:36.20,0:24:40.26,Default,,0000,0000,0000,,better and nicer if you could do it not\Ninvasively and you probably can if you
Dialogue: 0,0:24:40.26,0:24:43.60,Default,,0000,0000,0000,,improve the time resolution of EIT.\Nthere's nothing stopping you from doing
Dialogue: 0,0:24:43.60,0:24:50.36,Default,,0000,0000,0000,,that by the way. You just have to, like,\Nit's just a next step really.
Dialogue: 0,0:24:50.36,0:24:56.92,Default,,0000,0000,0000,,And then I'll also mention write-\Nfunctionality. So there was a paper that
Dialogue: 0,0:24:56.92,0:25:02.70,Default,,0000,0000,0000,,came out halfway through this year by a\Nguy called Neil Grossman (?) and what he
Dialogue: 0,0:25:02.70,0:25:09.17,Default,,0000,0000,0000,,did is, he showed that you can stimulate\Nneurons by sending current through the
Dialogue: 0,0:25:09.17,0:25:18.74,Default,,0000,0000,0000,,skull and in a focused way. Now why that's\Ninteresting is, you can non-invasively
Dialogue: 0,0:25:18.74,0:25:23.19,Default,,0000,0000,0000,,stimulate neurons. So that's the write-\Nfunctionality. It's unknown what
Dialogue: 0,0:25:23.19,0:25:27.95,Default,,0000,0000,0000,,resolution is or how well you could\Ncontrol the the focal point here. But it
Dialogue: 0,0:25:27.95,0:25:34.22,Default,,0000,0000,0000,,works in the principle of beat frequencies\Nso he sent through two kilohertz and 2.05
Dialogue: 0,0:25:34.22,0:25:42.38,Default,,0000,0000,0000,,kilohertz and basically had a beat\Nfrequency of 10 Hertz arise from that and
Dialogue: 0,0:25:42.38,0:25:50.28,Default,,0000,0000,0000,,basically stimulated neurons in this area\Nthat he can control via an x- and y-axis
Dialogue: 0,0:25:50.28,0:25:59.94,Default,,0000,0000,0000,,which is very impressive. Leaves a lot of\Nquestions open. Those are some possible
Dialogue: 0,0:25:59.94,0:26:06.31,Default,,0000,0000,0000,,next steps that it could go in. Obviously\NI think this is interesting. I hope that
Dialogue: 0,0:26:06.31,0:26:11.34,Default,,0000,0000,0000,,you do too. I'd love it if you would want\Nto sign up to a mailing list I'll give a
Dialogue: 0,0:26:11.34,0:26:17.05,Default,,0000,0000,0000,,link on the next page. If you want to\Ncollaborate email me. If you know any
Dialogue: 0,0:26:17.05,0:26:22.07,Default,,0000,0000,0000,,funding bodies that might be interested in\Nthe developing medical imaging for
Dialogue: 0,0:26:22.07,0:26:26.33,Default,,0000,0000,0000,,the third world I'd love to be put in\Ncontact. If you wanted a kit and, if there
Dialogue: 0,0:26:26.33,0:26:30.23,Default,,0000,0000,0000,,were enough people that wanted a kit,\Nprobably of the next version which would
Dialogue: 0,0:26:30.23,0:26:36.20,Default,,0000,0000,0000,,have 32 electrodes sign up to the mailing\Nlist, talk to me. Thanks.
Dialogue: 0,0:26:36.20,0:26:46.82,Default,,0000,0000,0000,,{\i1}applause{\i0}\NRintoul: Thank you
Dialogue: 0,0:26:46.82,0:26:49.69,Default,,0000,0000,0000,,{\i1}applause{\i0} Herald Angel: Thank you\Nvery much. We have a little bit
Dialogue: 0,0:26:49.69,0:26:57.61,Default,,0000,0000,0000,,time for Q&A. And please if you have to\Nleave the room make it in a very quiet
Dialogue: 0,0:26:57.61,0:27:05.73,Default,,0000,0000,0000,,way. So is there ... there are some\Nquestions I've seen microphone 4 first.
Dialogue: 0,0:27:05.73,0:27:08.97,Default,,0000,0000,0000,,Please go ahead.\NAudience member: So, a great thing
Dialogue: 0,0:27:08.97,0:27:15.65,Default,,0000,0000,0000,,thinking about developing countries and\Ngetting them medical tech. But at the very
Dialogue: 0,0:27:15.65,0:27:20.89,Default,,0000,0000,0000,,first beginning you said imagine a world\Nwhere this imaging would be all available
Dialogue: 0,0:27:20.89,0:27:26.61,Default,,0000,0000,0000,,like every day and it creeped me out a\Nlittle bit. Do you really think that it's
Dialogue: 0,0:27:26.61,0:27:33.74,Default,,0000,0000,0000,,a good idea to go in the shower in the\Nmorning and have your I don't know your
Dialogue: 0,0:27:33.74,0:27:40.02,Default,,0000,0000,0000,,bathtub telling you that there is a small\Nmass inside your lungs.
Dialogue: 0,0:27:40.02,0:27:46.89,Default,,0000,0000,0000,,Rintoul: That's a good question. Basically\Nthe question was: There's a privacy
Dialogue: 0,0:27:46.89,0:27:52.36,Default,,0000,0000,0000,,concern with looking inside your body. It\Ndoesn't sound that great to some people.
Dialogue: 0,0:27:52.36,0:27:56.20,Default,,0000,0000,0000,,To those people I would say you should\Nturn off I know that sounds a little
Dialogue: 0,0:27:56.20,0:28:05.18,Default,,0000,0000,0000,,harsh. But please just turn it off, don't\Nuse it. And with all scientific movements
Dialogue: 0,0:28:05.18,0:28:12.08,Default,,0000,0000,0000,,forward comes great risk, I also say. And\Nit can be used for good or evil and it's
Dialogue: 0,0:28:12.08,0:28:17.77,Default,,0000,0000,0000,,up to us as a society how we want to\Nchoose to use it. And how we structure
Dialogue: 0,0:28:17.77,0:28:24.51,Default,,0000,0000,0000,,ourselves and potentially motivate and\Nincentivize corporations to use it in a
Dialogue: 0,0:28:24.51,0:28:31.92,Default,,0000,0000,0000,,responsible way. Part of making this open\Nis I hope that, basically if people have
Dialogue: 0,0:28:31.92,0:28:36.47,Default,,0000,0000,0000,,access to it you can choose for yourself\Nhow you'd want to use it.
Dialogue: 0,0:28:36.47,0:28:40.81,Default,,0000,0000,0000,,Herald Angel: And next question would be\Nfrom the Signal Angel please.
Dialogue: 0,0:28:40.81,0:28:45.07,Default,,0000,0000,0000,,Signal Angel: Yes I have a couple of\Nquestions from the internet. First of all,
Dialogue: 0,0:28:45.07,0:28:51.44,Default,,0000,0000,0000,,what type of AC frequencies in use? the\Nasker assumes sinusoidal but he wonders if
Dialogue: 0,0:28:51.44,0:28:54.78,Default,,0000,0000,0000,,you also tried square wave, triangular and\Nother shapes.
Dialogue: 0,0:28:54.78,0:29:00.36,Default,,0000,0000,0000,,Rintoul: That's also a really interesting\Nquestion. It's about what kinds of waves
Dialogue: 0,0:29:00.36,0:29:08.58,Default,,0000,0000,0000,,are used, what kinds of AC signals.\NTypically it's done with AC sine waves
Dialogue: 0,0:29:08.58,0:29:14.50,Default,,0000,0000,0000,,ranging all over the place, depending on\Nwhat application you want to use up for. I
Dialogue: 0,0:29:14.50,0:29:19.98,Default,,0000,0000,0000,,mentioned multi frequency EIT for cancer\Ndetection. That uses a lot of different
Dialogue: 0,0:29:19.98,0:29:26.09,Default,,0000,0000,0000,,frequencies so if you wanted to use other\Nwaveforms I think that would be really
Dialogue: 0,0:29:26.09,0:29:32.84,Default,,0000,0000,0000,,interesting. Nobody's tried, you can, that\Nshould be done.
Dialogue: 0,0:29:32.84,0:29:38.74,Default,,0000,0000,0000,,Herald: So since there's a big queue on\Nmicrophone 3 I would go there please.
Dialogue: 0,0:29:38.74,0:29:44.63,Default,,0000,0000,0000,,Audience member: Yes I have a technical\Nquestion. Assuming that you won't use this
Dialogue: 0,0:29:44.63,0:29:50.59,Default,,0000,0000,0000,,techniques on humans or organic matter at\Nall and what are the limitations for the
Dialogue: 0,0:29:50.59,0:29:56.19,Default,,0000,0000,0000,,resolution. The spatial resolution. And is\Nthere a possibility to reduce the spatial
Dialogue: 0,0:29:56.19,0:29:59.10,Default,,0000,0000,0000,,resolution.\NRintoul: You mean increase the spatial
Dialogue: 0,0:29:59.10,0:30:06.18,Default,,0000,0000,0000,,resolution or reduce it?\NAudience member: Reduce the voxel size
Dialogue: 0,0:30:06.18,0:30:12.47,Default,,0000,0000,0000,,Rintoul: So increase the spatial\Nresolution. Yes absolutely. So I was
Dialogue: 0,0:30:12.47,0:30:16.24,Default,,0000,0000,0000,,trying to go through a few of the next\Nsteps that could get to that. One of them
Dialogue: 0,0:30:16.24,0:30:21.26,Default,,0000,0000,0000,,is magneto-acousto electrical tomography\Nbecause you get two different types of
Dialogue: 0,0:30:21.26,0:30:27.58,Default,,0000,0000,0000,,information which you could put together\Nto form a higher resolution image. So
Dialogue: 0,0:30:27.58,0:30:33.32,Default,,0000,0000,0000,,that's one way and if you didn't need to\Nworry about human safety I recommend you
Dialogue: 0,0:30:33.32,0:30:39.24,Default,,0000,0000,0000,,just turn the power up, that will also\Nwork.
Dialogue: 0,0:30:39.24,0:30:46.39,Default,,0000,0000,0000,,Herald: Okay I think we go back to the\Nsignal angel for one short one please.
Dialogue: 0,0:30:46.39,0:30:49.89,Default,,0000,0000,0000,,Signal Angel: Yes I have another question\Nfrom the internet. from a doctor this
Dialogue: 0,0:30:49.89,0:30:54.34,Default,,0000,0000,0000,,time. He wonders if there are any clinical\Nstudies that compare pulmonary edema
Dialogue: 0,0:30:54.34,0:30:59.92,Default,,0000,0000,0000,,diagnostics with EIT to ultrasound and why\Ndon't we just work on cheap ultrasound
Dialogue: 0,0:30:59.92,0:31:03.14,Default,,0000,0000,0000,,instead.\NRintoul: That's a good question. People
Dialogue: 0,0:31:03.14,0:31:08.29,Default,,0000,0000,0000,,are working on cheap ultrasounds.\NUltrasound gives different information to
Dialogue: 0,0:31:08.29,0:31:14.36,Default,,0000,0000,0000,,EIT. It has a problem of the sound\Nscattering. So it's a different type of
Dialogue: 0,0:31:14.36,0:31:21.29,Default,,0000,0000,0000,,information which has different pros and\Ncons. And and I think people should make
Dialogue: 0,0:31:21.29,0:31:27.17,Default,,0000,0000,0000,,cheap ultrasound. And I would like to see\Nthe hybrid modality come together. You can
Dialogue: 0,0:31:27.17,0:31:31.77,Default,,0000,0000,0000,,get really good tissue distinction with\NEIT so there's pros and cons.
Dialogue: 0,0:31:31.77,0:31:36.68,Default,,0000,0000,0000,,Herald: Okay then, microphone 2 please.\NAudience member: You had a really good
Dialogue: 0,0:31:36.68,0:31:44.65,Default,,0000,0000,0000,,talk my question so far you always need\Ndirect contact to the electrode, right? So
Dialogue: 0,0:31:44.65,0:31:50.52,Default,,0000,0000,0000,,it has to be direct contact or in water.\NIs there way to detect or measure the
Dialogue: 0,0:31:50.52,0:31:56.66,Default,,0000,0000,0000,,signal without direct contact? So maybe in\Nif the if the object is in air or any
Dialogue: 0,0:31:56.66,0:32:01.42,Default,,0000,0000,0000,,other gas?\NRintoul: Right. I wish there was. No is
Dialogue: 0,0:32:01.42,0:32:07.57,Default,,0000,0000,0000,,the short answer. Unless ...\NAudience member: Any research on making it
Dialogue: 0,0:32:07.57,0:32:11.13,Default,,0000,0000,0000,,happen?\NRintoul: Well yeah you can you can use
Dialogue: 0,0:32:11.13,0:32:20.13,Default,,0000,0000,0000,,X-rays. They work wonderfully to to go\Nthrough the air. But if you use them I
Dialogue: 0,0:32:20.13,0:32:24.22,Default,,0000,0000,0000,,mean you do increase your chance of cancer\Nso don't use them all the time on
Dialogue: 0,0:32:24.22,0:32:29.50,Default,,0000,0000,0000,,yourself. Again CAT scanners are a little\Nbit expensive.
Dialogue: 0,0:32:29.50,0:32:35.53,Default,,0000,0000,0000,,Herald: Thank you and I think we have time\Nfor one more from microphone 3
Dialogue: 0,0:32:35.53,0:32:41.97,Default,,0000,0000,0000,,Audience member: My question would be\Nwhat, so maybe I've missed it, but what's
Dialogue: 0,0:32:41.97,0:32:47.01,Default,,0000,0000,0000,,the order of magnitude for cost so would\Nthis be feasible at like a hackerspace for
Dialogue: 0,0:32:47.01,0:32:53.75,Default,,0000,0000,0000,,this to implement. And does the industry\Nsee the possibility to make money.
Dialogue: 0,0:32:53.75,0:33:00.78,Default,,0000,0000,0000,,Rintoul: Yes a lot of those sort of these\Nearly like R&D papers yeah they should be
Dialogue: 0,0:33:00.78,0:33:06.87,Default,,0000,0000,0000,,applied and you could make money with it\Nabsolutely. And there's no component in
Dialogue: 0,0:33:06.87,0:33:14.96,Default,,0000,0000,0000,,there that costs more than a couple of\Ncents. I suppose a cortex m3 like costs a
Dialogue: 0,0:33:14.96,0:33:19.57,Default,,0000,0000,0000,,couple of dollars. And I mean I don't know\Nwhat your budget is but yes you I think
Dialogue: 0,0:33:19.57,0:33:24.31,Default,,0000,0000,0000,,you could do this in a hackerspace without\Nany problems. There's nothing stopping
Dialogue: 0,0:33:24.31,0:33:29.59,Default,,0000,0000,0000,,anyone from doing this and as we know\Nmicrocontrollers are becoming cheaper and
Dialogue: 0,0:33:29.59,0:33:36.46,Default,,0000,0000,0000,,cheaper. So why not.\NHerald: I don't get Hasty's signs from the
Dialogue: 0,0:33:36.46,0:33:40.06,Default,,0000,0000,0000,,sideline so I think I can take another\Nquestion from 2 please.
Dialogue: 0,0:33:40.06,0:33:46.60,Default,,0000,0000,0000,,Audience member: So far you have showed us\Nimages of 2d planes. What about volumes
Dialogue: 0,0:33:46.60,0:33:52.37,Default,,0000,0000,0000,,Rintoul: Yes so there's work on solving\Nfor volumes using finite element models
Dialogue: 0,0:33:52.37,0:34:02.76,Default,,0000,0000,0000,,and solving Maxwell's equations. Basically\NI just did the shortest route to reach
Dialogue: 0,0:34:02.76,0:34:07.67,Default,,0000,0000,0000,,image reconstruction that was available\Nwhich was linear back projection which is
Dialogue: 0,0:34:07.67,0:34:12.28,Default,,0000,0000,0000,,typically done in a 2d plane. So\Nabsolutely, you can do it in three
Dialogue: 0,0:34:12.28,0:34:16.37,Default,,0000,0000,0000,,dimensions.\NHerald: So I'm very sorry we are out of
Dialogue: 0,0:34:16.37,0:34:23.55,Default,,0000,0000,0000,,time the queue back there you can have the\Nchance to chat with our speaker just right
Dialogue: 0,0:34:23.55,0:34:31.88,Default,,0000,0000,0000,,now. The next talk coming up is in about\N15 minutes and it's I think also in
Dialogue: 0,0:34:31.88,0:34:36.92,Default,,0000,0000,0000,,English. See you then and a big round of\Napplause for our speaker, excuse me.
Dialogue: 0,0:34:36.92,0:34:42.46,Default,,0000,0000,0000,,{\i1}applause{\i0}
Dialogue: 0,0:34:42.46,0:34:47.72,Default,,0000,0000,0000,,{\i1}music{\i0}
Dialogue: 0,0:34:47.72,0:35:04.00,Default,,0000,0000,0000,,subtitles created by c3subtitles.de\Nin the year 2017. Join, and help us!