24. Mai 202117 Min. Lesezeit

eye2you Converts Smartphones in Simple Medical Retina Scanners

New Blog

This blog post first appeared first on old medium publication (https://medium.com/startuprad-io), and was moved to this blog with the relaunch of our website in summer 2024.

Tune in to our Internet Radio Station here:

www.startup.radio

Subscribe Here

Find all options to subscribe to our podcast, listen to our internet radio station, or follow us on social media here:

linktr.ee

The Founder

We are talking to Jörn-Philipp Lies (https://www.linkedin.com/in/plies/) CEO and Co-Founder of eye2you (https://eye2you.ai/).

Philipp started out his professional life with an apprenticeship and moved to study at Technical University Darmstadt (TU Darmstadt, https://en.wikipedia.org/wiki/Technische_Universit%C3%A4t_Darmstadt) in computer science and robotics.

While working on his final diploma thesis, he became interested in visual recognition and finally the eye. This led him to take on his Ph.D. at the University of Tübingen, studying how the brain processes visual information.

The Startup

The story of eye2you (https://eye2you.ai/) started at Korea University (https://en.wikipedia.org/wiki/Korea_University), with a project for development aid projects in Ethiopia. Now the company is incorporated in Tübingen Germany and works on AI algorithms to recognize diseases in pictures of the retina.

They are working on empowering general physicians, nurses, or caretakers to perform a basic checkups with their software. Groups at risk for retina diseases are people with high blood pressure or diabetes.

Here is an explainer video for more details:

Venture Capital Funding

The startup already closed the pre-seed funding round, currently having 5 FTEs and one working student.

The Video Interview is set to go live on Tuesday, May 25th, 2021 at 17.00 CET

The Audio Interview Podcast

The Audio Podcast will be published at the same time. Please give us a bit to post the Spotify link here. In the meantime, you can subscribe to our podcasts here.

Feedback

Reach out to us, here is our audience survey, to give us feedback, suggest topics, interview partners or just to say “Hallo!” https://forms.gle/mLV6mVKwGwKuut8BA

The Interviewer

This interview was conducted by Jörn “Joe” Menninger, startup scout, founder, and host of Startuprad.io. Reach out to him:

Twitter

Transcript

[0:00] Music.

[0:08] That I owe you were podcast and YouTube blog covering the German.

[0:14] With News interviews and Live Events hello and welcome everybody this is Joe from celebrated oh your startup podcast and you to block from Germany as well as to what’s first,internet radio station dedicated to startups and Tech.Companies today I bring you another startup here from lovely Germany officially headquartered in tubingen and Philip as my guest hey Phillip how you doing hi Joe I should have me to be on the show.

[0:47] Yeah it’s totally pleasure to have you here we already talked beforeyou try to unfix the the bicycle behind you so it will fall on your head during the interview and we have nice take out rightthat was the agreement exactly yeah Lucent the screens so everything is set up to be like six to seven minutes will fall right in my neck and then we have a very nice I’ll take.

[1:10] In order not to miss the outtakes or anything else make sure to hit the like And subscribe button wherever you watching this or listening to this if we are referring to something like Phillips.Um LinkedIn profile the universities or the startup make it go down here in the show notes there’s always a link to our blog post where you can learn more that set I.Was a little bit surprised when I went to your CV because officially you guys aremed tech startups but then I realize huh you’re not a medical doctor you are a computer scientist neuroscience.PhD but first let us talk a little bit aboutwhat you did because I’ve seen you studied computer science at Technical university darmstadt university from which we already had a lot of alumni here in our interviews can you take us a little bit likethrough your life what you’ve done so far and how it let you two found eye2you.Yeah I’m very happy to so before I started my my studies in darmstadt I did an apprenticeship as a chemical lab assistant.

[2:31] After that decided okay I want to do some more Tech stuff so I went to to darmstadt University,which is one of the best universities for computer science they have a very large,variety of topics and I was interested always interested in what do how do robots do how does vision work how do you visualize things.And that’s why I decided okay da rooster is the perfect place for me,so I’m going to darmstadt and did my diploma back then there will focus on robotics computer vision and 3D graphics.And my diploma thesis 10 was about developing a camera system for robots that helps detecting Things based on how you use them and not what they are so instead of saying this is a cup this is a,Jack this is a plate,the camera system was to take the okay this is something I can grasp at the handle this is something I need to grasp on the side and saying this is more are more natural way because you don’t need to categorize the objects it’s more about how to use them,that sounds pretty trivial at the beginning but I’m very sure it gives you gave you a lot of headaches when you try to teach that to computer yeah.

[3:47] Yeah it was a very very exciting topic and we were working on an international team with people from Birmingham and from from Italy and it was very very nice project and I was a lot of fun back then.

[4:00] But then I decided okay I spend weeks and weeks on fiddling around in Minor Details In camera systems for robots,and said okay but we have very very good camera systems natural camera systems,so I want to understand how do they work so how does vision work how is it possible that I’m standing here I can see you you’re sitting there you can see me you can see my background the bike you know I’m in the front of the bike is in the back all this kind of processing goesbecause in your head so how does this work so I was looking around okay where is a good place to do this.And then I found Professor Bitcoin in tubing at the Max Planck Institute for biological cybernetics.And he had their junior research group they are back then and I was doing very very exciting research in,computational neuroscience and said that this is exactly what I wanted to do,so I wrote him an email and explain what I did before and what I want to do now and I was asking him for a for a PhD position and luckily he already did you just had a PhD position open for somebody with my my track record.SoStarted talking to him and me and then we decided okay sounds like a good match so I went to tune him and yeah started my academic career then into being.

[5:17] We may ask for everybody who’s not a very frequent follow of this podcast.Or not familar with the Germany the Max Planck Society is like an association of very research heavy institutionsand it’s named after the German researcher Max Planck who discovered.

[5:42] I think he laid the groundwork for quantum theory with the write something like this I’m sorry guys I’m not a physicistI’m sorry that that just my rude understanding if you do have anything else to add make sure to do to teach us down here in the commentsand so basically that as we said it’s a research heavy institution so what was your research there.

[6:12] So my research dad was on how nerve cells in the brain process the information.So when you get when an animal human is getting information you have the retina in your eye this takes the light and then transfers it into electrical signals that goes through your nerves.And then the sequence bill from from the the eyes,through the through the back of the head and start processing there and then go into different levels and in your in your brain.

[6:42] Met at a very simplified version basically you.

[6:49] You were working how to decode the human graphics card.

[6:54] Yeah so it was okay how do these these cells in the brain work how do they process information,we’re kind of optimization do they use so understanding how these this little cells work,and this is older than like the bases that’s going on with all the artificial intelligence systems we have now so all these like deep learning topics where you have.Like deep fake is a very prominent topic now where you have all these these methods in,with real networks that can generate images that can fake make fake images and can generate Texas and everything,so all this is some part of what I was doing my my PhD which I did into being a for about five years first at the Max Planck Institute,and then my my professor went from the Max Planck Institute to the University Hospital in 2 million.

[7:46] And then the second part of my PhD at the University and the University Hospital in tuebingen,we may also add that the University of tubing is one of the Premier Researchuniversities in Germany which is talk before and for example DNAwas discovered there the basics of DNA and some famous alumni like young stock the formerChief Economist of the Year Penn Central Bank radiated from there as well as lot ofvery famous people in theology psychology geography and so on and so forth so it they have like hundreds of famous alumni in certain Fields so you’ve been there you studied how the brainprocesses visual information and this let you.

[8:43] In what way to start an app.

[8:48] That cares about the I how how did this how did this twist this turn actually.

[8:56] Actually there is one step in between which we are currently missing so after I finished my PhD.I started working for German consulting company it and.Strategy Consulting for different blue-chip companies in Germany.And I worked there for about five years so like didn’t get in automotive industry in insurance industry did like project management for them all these agile transformation and everything,and after about five years I said okay this is all nice I learned a lot it was great experience,but I wanted to go back to where I came from to tack to be closer to research.And then it was like I just put my job and was then about okay what I want to do something something more meaningful with with my time.

[9:47] And so I was checking around with like friends what they’re currently doing where they’re working and then I was talking to a friend of mine Christian who is a professor at Korea University in Seoul.And was asking him so here don’t you have anything like a project where you could use somebody who has experience in like all these artificial intelligence machine learning topics and he said yeah,I have a PhD student here he’s working in developmental Aid in his free time.And he’s doing a project in Ethiopia where they record images of the retina because this PhD student is an.

[10:25] And Ultimate Edition so he knows how to record the retinas and say yet they’re really good at recording retinas they have the equipment but they ever have very little opthamologist who can then take these images of the retina and then say okay,this is a healthy retina this is has some kind of disease.So this could be something where we apply machine learning to help the people there to make this faster and more easily.And Sergei the sounds like a good idea so I just like.Book the next flight to to Seoul and then went to Korea to Seoul and started working on this research project there which was about two and a half years ago.And at first we thought okay let’s try something out for the branch of Developmental Aid let’s see if we candevelop an algorithm that’s like very small and easy to use so we can give it to small hospitals in Ethiopia so they can use it there,and then we follow him this this kind of works and.

[11:26] Played around with like a how can we do this on smartphones for example because it’s a devices that they available everywhere.Okay this is actually a very nice idea and then we started talking to two people back in Germany,and this area actually this is a really nice idea so this could be a really useful tool not even in Africa but also in other countries like in Germany all over Europe.That’s okay then let’s let’s see if we can this idea that we came up with in or a couple of beers in a research project because when we turn this into a company.

[11:57] So I started talking to Philip Barons who is a professor at the Eye Hospital in tubingen and said hey this is our idea and we want to turn this into a startup company.

[12:09] Do you want to join us this is a nice idea where you can work together,and he said yeah actually the University Hospital tubing is very interested in continuing working on this this kind of research and forming into a companies so we applied for start-up funding in Germanyand they were also really excited about the idea and that’s why we got then startup funding in the beginning of 2020.It is an interesting idea basically,you went from darmstadt to tubing into Korea just to get back to tuebingen that at that at that is quite an interesting Journey when you’ve been talking about likerecognizing diseases in your eye that’s basically what we’re talking about I hadtwo thoughts first could you like also integrate this in a smart mirror for example you you look in the mirror every morning and just.By the snap of a finger basically every morning you’re smart mirror checks your eyes is that possible and secondly is it also possible to diagnose,other diseases not of the eye when you’re scanning the eye before we get into the nitty-gritty details of what you guys are actually doing.

[13:24] This will actually be a really great application to have business smart mirrored the problem is to see the retina so the back of the eye you have to go through the pupil and then see the back of the eye,and for this you need some kind of lenses so you need some kind of Optics so the very simple one is something like this here,so this is a mobile phone the scope where you can put in a camera in the back and have then the the smartphone and then record the retina with this.So it’s not possible to it like have a mirror and then just have a camera chip there which these through this very tiny black spot in my eye.So you need some kind of object to put in front of the eye and record with the retina.That’s why this like this mirror idea is actually great and might work for diseases of the outer part of the eye,but not for diseases of the retina.And for the other part of actually yes the retina is very interesting because it’s part of the of the central nervous system central nervous system.

[14:26] So you can see also black fine blood vessels and everything without having to penetrate the the body so you can see the very fine structure of the.Of the vessels of the blood vessels and this gives you a lot of information about other disease is to for example hypertension.So there is a lot of potential in using images of the retina for diagnosing different kinds of diseases so I think later we can go between detail on that.And now talking about your company we already realized you are using smartphones,choose skin the eye and you’ve shown for everybody who’s not watching this as a YouTube interview but listening to it as anaudio podcast you showed something you had to put on the camera of the smartphone which more or less looks like a little handgun and with this you can scan the I amcan you tell us a little bit more about what you guys actually doing what setup you need and where your app is currently.

[15:33] Um in use and how you guys are working with this and with this data.Yeah sure first part want me to do you say it looks for you like little handgun I think that’s your text and background speaking there because most people I talked to they say it looks a little it back like a hair dryer like a big hair dryer.But it’s so this actually it is a small device a handheld device the size of of the of the addres hair dryer.And this is more or less simply speaking just a couple of lenses that.Focus the light from from the eye and that goes into the eye to get the images back and what we are.

[16:15] Currently doing is we’re on an early stage so we’re currently in the development stage together with the Eye Hospital in tuebingen,where we developed a smartphone app that helps recording the images because you need to go through the pupil as it’s very small,small parts so you need some kind of assistance to use this this device to record the.Retina properly and then once you have a properly recorded retina this image then gets analyzed by different kind of algorithms to see okay is there.This kind of disease or that kind of disease.And the application we are going for is not that the Epic person of a customer just uses this device for themselves but we want to empower general physicians and caretakers with this kind of device,because we believe artificial intelligence especially in the medical field.Is something that still needs as like a last decision made from medical expert.

[17:16] So having this this method say okay you have this and this disease just without medical supervision I think can be kind of dangerous for misinterpretation.So that’s why our belief is we should use this as a smart tool to make doctors more powerful than they currently are.And make examinations more easily accessible.Currently if you want to get this kind of retina examination you have to go to an ophthalmologist office,which there are at least in Germany we have 80 million lot more than 80 billion people here and only 8,000 opthamologist in Germany.So to get to an opthamologist usually have to drive a long distance then you have to wait in the waiting room and if and the cost is a lot higher for the health insurances because experts are more expensive.

[18:07] And if you see okay how many people are at risk for retinal diseases which are people over a certain age which are diabetic people,or as I said before people with high blood pressure so this is a large part of the of the German population and worldwide population,diabetes is it’s a very dangerous disease all over the world.

[18:28] So there are a lot of people who might benefit from this like regular and easy checkups and we believe if this checkup can be done by your family doctor or bye,by a caretaker in a nursing home that is a lot more accessible and you can find retinal diseases lot easier and a lot faster.And this is why our approach is we want to empower these people with,the software we develop for the smartphone so they can use their smartphone or that the office smartphone together with these.Is the mobile fundus cameras which are regular medical devices you can can buy off the shelf.

[19:04] And then do this kind of examinations to see okay is there,some indication of a disease and if yes then they should go to the expert for the treatment so we don’t want to replace opthamologist and experts we just want to add an additional layer of checks.Between doing nothing and going through the expert because data is a lot of necessity in between.And that’s why we think this is a very good idea which is useful in.Highly technological countries like in Germany or in Korea all over Europe but also in other countries where you don’t have medical infrastructure like in developing countries in Africa let me try to.Poke a little bit on a software how good is this software andhow do I see the results as basically like a traffic light green everything’s fine orange you may develop something in red you see should see an opthamologistor is it something a little bit more dedicated,how high is your confidence in this software tool meaning does it like always get,D retinal diseases.And so the quality of these AR algorithms to detect for example diabetic retinopathy this is like the one of the first diseases that we are currently working on.

[20:29] Daddy how us firms are even on an expert of tumultuous level or even better so you’re in the 98 99 percent of.Correct detection which is more or less the same as as expert doctors can recognize these diseases.

[20:45] But we are actually going for this kind of traffic light approach to say okay,for caretakers and family doctors it’s not important to to see okay this is a stage 3 or stage for retinopathy for day of them it’s important okay,this is a this is a clinical case so this kind of person needs treatment for the retinopathy so this is then like a red case or okay there seems to be some first indication on retinopathy,maybe check with the General Health of the person so if there might be a good reason for a check-up or green say okay everything is fine.

[21:22] So this is the the recommendations that our our software gifts.But still has in the back all the information that if the if the doctor needs some more information,then he or she can get a full report on everything but the first location should be okay this is nothing or if there is some might be some case or this is a very secure case so this person needs immediate treatment.

[21:45] ICC and second question would be howdoes it proceed as AC said okay let’s assume I get this ice can’t there is like a red light andwhat happens then there is a caretaker and he says you should go see an eye doctor an opthamologist or do you actually enable the people to go there with the dataopthamologist to access the data do you work together with for example and Hospital who gets like if approved all the pictures how do you guys work then,so I’m currently so we’re slowly developing face so this is not not yet fully implemented but our current plan is that we that the recordings you do take,they go into the patient management system of the doctor’s office that you’re at so for example you’re at your family doctor,they take the image and then they get a full report just like foreign EKGs and store this in your patient file,and then we say okay there seems to be a case you should go to see an ophthalmologist,they can do the report and then send it EA either digitally or just do a print out and hand it to you which you can then bring to the ophthalmologist.But then the optimal just would do a full examination anyway because after Ma just have a lot,has a more high quality Equipment they have very big expensive stationary cameras.

[23:10] Which cost several hundred thousand Euros and then do with this kind of special camera the examination,because devices like these here they are in the their clothes cost a few hundred to up to like one or two thousand Euros so the quality of these is totally sufficient and very good for doing quick checkup.But if you want to have this exact distinguishing between okay this is a stage 3 where’s the stage for or maybe there’s some some other Minor Details,for this you need this this big a hundred thousand Euros cameras and this is what the opthamologist in their office app they do the recordings with this and then they can very finally decide okay what kind of treatment do you need.

[23:50] I do believe even the opthamologist like it because it only brings like people to their to their attention that actually,half a suspicion to be really sick they’re written are really sick and you already get an indication,where to look I do believe that and that’s a big plus especially if you also get in report and then you can crank up your expensive equipmentthat is pretty awesome let’s talk a little bit about your company how many guys are you right now and how are you financedSo currently we’re a team of five and a six mile company 5A F 5 + + working student,and yeah we were founded last year in the beginning of 2020,then because of the pandemic not everything was working as planned so the our growth was a little bit blocked by this.

[24:51] And now we are we finally closed our precede Finance Round so we currently proceed Finance with from the.Hearing from Bank in Germany.

[25:02] And we have now the founder team are the three founders Beyond Christian me all with a more technical background and then we have another members with a more medical business background in our team.And yeah this is the current state and we are still looking for more people and especially for the regulatory part here in Germany to get our software.Certified as a medical product according to the European medical device regulation.And some more assistance in different fields so for people interested in working with us can check out websites there we have some information on open position now.

[25:42] I see and everybody would like to learn more can go down here in the show notes we do have a lot of Wikipedia articleson Ritchie nope a fee opthamologist and so on and so forth and of course we also have your personal LinkedIn profile and people will realize I always call you Philip.Yearn Philip but we agreed just to call you Phillip and that’s totally fine and the people will also find your LinkedIn profile.Down here only thing left for me to say is.Thank you very much it was a pleasure having you as a guest and best of luck for the continued development.Of your tools they had like any any timeline in mind when you will be registered,as an official medical device medical software and yes our plan is to get the certification done by the end of 2020.So about the end of next year and.So this is the current plan to have them ready and started working in 2023 with the the first doctors most likely in South Western German part so starting locally,and then expanding to Germany Europe and then maybe in two or and three or four years also internationally.

[27:07] Well best of luck and hope to have an update with you guys as soon as we have.As soon as you are expanding internationally thank you Philip thanks Joel pleasure to be here.

[27:21] If you are a professional looking at the European startup scene Germany is a place you cannot miss.

[27:28] Music.

[27:38] Most likely you have never heard or read anything only startups before in English but you will in the future be ahead of the curve And subscribe to.

[27:48] Music.

STARTUPRAD.IO