Camels, Code & Lab Coats: How AI Is Advancing Science and Medicine

there’s been many innovations over the years that dried science if you go back in history many scientific insights actually derived from new tools that were able to measure new things AI is very very good at finding new paths that haven’t been seen before it’s almost an enhancement on our ability to sense the world the main benefit of machine learning is the ability to learn from lots and lots of data we can show a computer a lot of examples and rather than tell a model how to evaluate that data can learn actually how to interpret it and figure out what to do next here at Google we’ve been using machine learning technologies in all of our products things like search translate Google photos and the assistant people are realizing that learning from examples is a very powerful tool so researchers in the healthcare community we’re looking at some of the work that Google was doing in artificial intelligence and actually reached out to us and said is there a way that you could apply these same kind of technologies in healthcare doctors today have more artificial intelligence working for them on their smartphone for their personal use than they have working for them in the clinical context and as doctors we’ve just gotten more information kind of shoved at us in terms of Records in terms of images think about all the different types of selling things you have 15,000 20,000 different diagnosis codes just a cubic millimeter of tissue it’s like taking you know a billion photos if you print it out they would probably be about as tall as a ten story building really any kind of data that people can generate on a massive scale is really an area where machine learning could help now there’s an opportunity to use all our digital technologies that have been developed at Google to really try to help doctors let’s take some of these tools that are great for analyzing videos and YouTube and apply them to problems it matter to science one of the things we’ve been working on in medical imaging is in the area of pathology traditionally pathologists take some tissue sample and they look around and the see of the cells looking for the kind of needle and haystack cancerous tissue we know that the earlier you detect the cancer the greater your chance of carrying it and the greater your chance of curing it without chemotherapy or radiation so the biggest challenges are speed and I’d receive diagnosis so far we’ve trained models for breast cancer and prostate cancer these technologies can actually identify suspicious areas to direct the doctors attention so one of the things we wanted to do was get this work into the hands of as many people as possible and so we developed something called the augmented reality microscope where you can actually see machine learning assistants overlaid in real time as you’re looking through the microscope these units can be attached onto any existing microscope greatly reducing the cost we’re really excited to bring machine learning to parts of the world with limited access [Music] one of the tools some biologists use as they will dye cells with different colors to highlight certain important features that make sense to them the problem is you have to kill the cells in order to color them so we said well we can do this virtually in a computer we can preserve the cell in its natural state one of the things that you can do on your pixel phone right now is given a selfie you could predict the depth and you can do interesting visual effects so we thought that hey can we take the same technology and apply it in a biological context essentially we use machine learning to predict a staining we weren’t sure whether this would work or not and it ended up working so well we’re very hopeful that this technology can be used to just have the computer or generate the pretty pictures that people know how to interpret the brain is probably the most complex physical object in the known universe we know that there are these basic units called neurons and they’re connected in many different ways what’s shocking is how little is known about what those patterns of connectivity look like and what that means for how the brain works the problem is very hard because the connections are too large to analyze for example fly brain has a hundred thousand neurons whereas a human brain has 100 billion neurons so fortunately at Google you know there’s been already a lot of work put into dealing with datasets of that size machine learning and the computer vision technology that we’ve developed has been designed to accurately trace the wiring of the brain in 3d prior to that technology would have taken thousands of years to basically finish mapping the fly brain now you can do it within a year or two it’s a warm-up to understanding larger and more complex brains hopefully human brains we’re hoping that mapping a brain could potentially help us understand a lot of the neurodegenerative disorders the for example schizophrenia or Parkinson’s then we’re going to be able to design better therapies that might improve those conditions they’re tremendous inequities in the way that healthcare is distributed across the globe any disease or outcome is predicted as much by your zip code as it is by your biology so what can we do with AI to bring the expertise to where no expertise exists one of the complications of diabetes is diabetic retinopathy causes blindness and it’s diagnosed by seeing little lesions in the eye but in India there is a shortage of eye doctors and as a result about half the patients suffer some form of vision loss this disease is completely preventable this shouldn’t really be happening so we were able to train a model to reap these images and match board-certified ophthalmologist we’re now figuring out how to deploy this into the clinic in India people who did not have access now have access a lot of signs are driven by hypothesis that someone has but some of the biggest breakthroughs and science come from surprises things no one expected to have happened one of the really exciting things about deep learning is where you could just give it the raw data and it finds the important features that’s interacting with scientists at a very different level and saying well let me show you something you’ve never seen before we had one research project we were looking at human retinas but what surprised us was machine learning started seeing things that people didn’t know was possible turns out that a deep learning model is actually able to identify things have nothing to do with your eyes like your cardiovascular health your metabolic profile these are things that if you had asked experts what do you think we’ll find they would have said nothing the important thing about this is it’s a visual biomarker that we did not know existed before those unexpected things may lead to a whole new idea a whole new approach a whole new hypothesis for how to attack the problem you’re trying to address there are real patients suffering today and if we’re not doing everything we can with all the technology or disposal to help those patients then what are we doing I think that 10 or 15 years from now the use of machine learning in healthcare is just going to be how healthcare is now I think this is how we’re gonna find new discoveries I think this is how we’re gonna find ways to take care of more people more we can accelerate that basic work and give all these Sciences new tools we will as as humans really benefit from the new discoveries that when they end up in our doctors offices [Music]

88 comments

  1. เน„เธกเนˆเธฃเธนเน‰

  2. I wish Google would develop consumer oriented hardware to support their awesome software… A biometric-centered smartwatch maybe ?
    If I'm defaulting to an Apple watch I'd be forced to also use an iPhone, and I'd like to avoid that…I really would.

  3. come on Google, It's an alien technology!! it's beyond human technology and capabilities. This technology is not possible without outsiders.

  4. If we're not doing everything we can with all the tech at our disposal to help those patient, then *what are we doing*.

  5. ์ž๋ง‰์„ ํ•œ๊ตญ์–ด๋กœ ๋ฐ”๊พธ๋ฉด ์ •๋ง ๋งŽ์€ ๋„์›€์ด ๋ฉ๋‹ˆ๋‹ค ๊ทธ๋Ÿผ ์–ด๋Š๋‚˜๋ผ์—์„œ

  6. Love this ๐Ÿ˜๐Ÿ˜˜ I'm a computer science student. I like to learn about big data and machine learning ๐Ÿ˜๐Ÿ˜๐Ÿ˜

  7. I wanted to work with MRI models to train an AI project that I had in mind but the hospitals I consulted had privacy concerns about my idea ๐Ÿ™
    I just wanted to discover ways to detect no threatening scans in ms in order to clear the already packed waiting rooms in ERs

  8. so what you're saying is you can use advanced AI to be a personal fact hording device that when asked a question can repeat something it's already learned? and connect to the internet to learn how to smelt metal? based on google searches of what copper ore looks like and how to smelt it?

  9. People have been thinking about these things for a long time, eg. caduceus system from the 1980s, yet the integration of AI into today's diagnosis medicine feels decades behind what it should be. I think you guys should approach the medical system in smaller countries and trade computing power and AI assistance for raw data. Like give every GP office in that country free access to an AI platform where they can log all patient information (under anonymity of course) and where the software displays possible diagnosis, further recommended tests, prognosis and maybe hidden links for epidemiologists, etc. Patients get better treatment overall, the doctors get a lot of help the they need and you guys get a huge data gold mine. Positive sum-game for everyone involved imo.

  10. Electromagnetic spectrum filtering. See invisible light. See sounds. Artificial taste, smell, touch. You need more sensors. Artificial nose. Artificial ear. Artificial hawk eye, fish eye, dog eye. Pressure sensors. Filter out within a range. Copy nature, artificial bats ear. Sense combinations. Black and white picture and assign colour to sounds, xrays, radiobwaves. Fractal growth. Artificial cells. Artificial organs. Multiple world views, Multiple analysis. 3d printed damage detection nano cells, Micro sensors and cameras. Create a human to machine interface, create a brain cloud, we need a brain app store like play store for brain apps, like you should be able to download a specific set of skills.

  11. What should I do to get a job of R&D in that thing?
    I am a student and I am interested in both computer science (specifically AI) and biology. I want to pursue a career in this. Any advice please ๐Ÿ™‚

  12. EARTH'S FIRST SMARTPHONE DIAGNOSIS DOCTOR…๐Ÿ‘๏ธโ€๐Ÿ—จ๏ธ๐Ÿ‘‚๐Ÿ“ฝ๏ธ๐ŸŒŽ๐ŸŒ๐ŸŒ๐ŸŒ=โš•๏ธ

  13. This completely reminds me of Baymax!! I just can't wait to actually strive hard to work for those who need me and assisstghem to my fullest capability!

    Ooh.. I sound like I'm trying to impress my interviewer or something.. but who cares? I really wanna do something good with my life, and Google just inspires me more and more!

  14. You guys were at SIGGRAPH. Did any of you happen to see Quantitative Imaging's talk on the intersection of graphics and medicine? You guys should talk with them, maybe you could help.

  15. Sac๐Ÿ’€โ˜ ๏ธโ˜ ๏ธโ˜ ๏ธโ˜ ๏ธ๐Ÿ’€๐ŸŽƒ๐ŸŽƒ๐Ÿ’€๐Ÿ’€๐Ÿ˜ก๐Ÿ˜ก๐Ÿ’€๐Ÿ˜Ž๐Ÿ’€๐Ÿคก๐Ÿ˜ก๐Ÿ’€๐Ÿ™‹๐Ÿฝโ€โ™€๏ธ๐Ÿคก๐Ÿคก๐Ÿ™‹๐Ÿฝโ€โ™€๏ธ๐Ÿคก๐Ÿ‘พ๐Ÿ‘พ๐Ÿ‘พ๐Ÿ‘พ๐Ÿ‘พ๐Ÿ€๐Ÿ€๐Ÿฟ๐Ÿ“๐Ÿˆ๐Ÿ๐Ÿ๐Ÿ‘๐Ÿฉ๐Ÿˆ๐Ÿˆ๐Ÿˆ๐Ÿˆ๐Ÿ๐Ÿˆ๐Ÿ’€๐ŸŽƒ๐ŸŽƒ๐Ÿ‘น๐Ÿ‘น๐Ÿ‘น๐Ÿ’€๐Ÿ’€๐Ÿ’€๐Ÿ’€โ˜ ๏ธโ˜ ๏ธ๐Ÿ‘พ๐Ÿ’ฉ๐Ÿ‘พ๐Ÿ’ฉ๐Ÿคก๐Ÿคก๐Ÿ’ฉ๐Ÿ’ฉ๐Ÿ’ฉ๐Ÿง˜๐Ÿฟโ€โ™€๏ธ๐Ÿ‘ป

  16. Google I love you so much, I can't survive without you.
    Thank you so much for being there for me.
    Everyday I learn something new, congratulations ๐Ÿ’•๐Ÿ’• you make a difference .

  17. F๐Ÿคฏ๐Ÿคฌ๐Ÿคฅ๐Ÿคฅ๐Ÿคฅ๐Ÿคซ๐Ÿคฏ๐Ÿคฌ๐Ÿคง๐Ÿค—๐Ÿคข๐Ÿค—๐Ÿคง๐Ÿคฎ๐Ÿคฎ๐Ÿ˜˜๐Ÿคฅ๐Ÿ˜ต๐Ÿ˜ต๐Ÿ˜ตโ˜ƒ๏ธโ˜ƒ๏ธ๐ŸŸ๐Ÿ—๐Ÿ—๐Ÿ—๐ŸŸ๐Ÿ‘จโ€๐ŸŽ“โ˜ƒ๏ธโ˜ƒ๏ธ๐Ÿ˜Š๐Ÿ˜Š๐Ÿ˜Œ๐Ÿ™ƒ๐Ÿ˜˜๐Ÿ˜๐Ÿคฏ๐Ÿง๐Ÿ˜ž๐Ÿคช๐Ÿคฉ๐Ÿคข๐Ÿคง๐Ÿคฎ๐Ÿคข๐Ÿคข๐Ÿคข๐Ÿ˜ท๐Ÿ˜ช๐Ÿค‘

  18. Thank you Google. I've been living with type 1 diabetes since I was two years old. I'd say this is just in time for me.

  19. kayce
    wrd๐Ÿ˜‚๐Ÿ˜…๐Ÿ˜…โ˜บโ˜บ๐Ÿ˜Š๐Ÿ˜Š๐Ÿ˜Š๐Ÿ˜Š๐Ÿ˜Š๐Ÿ˜Š๐Ÿ˜Š๐Ÿ˜Š๐Ÿ˜…๐Ÿ˜…๐Ÿ˜…๐Ÿ˜…๐Ÿ˜…๐Ÿ˜…๐Ÿ˜…๐Ÿ˜…๐Ÿ˜…๐Ÿ˜…๐Ÿ˜…๐Ÿ˜…๐Ÿ˜…๐Ÿ˜…๐Ÿ˜‡

  20. Awesome work!! Only point i dont agree with is at 6:45. The study of the eyes, Iridology, has done this for a long time. Would be great to see this technology used to expand the studies into the eyes and look forward to what else can be achieved ๐Ÿ™‚

Leave a Reply

(*) Required, Your email will not be published