Remote Diagnostic Testing and the Machine Learning Revolution

On Tech & Vision with Dr. Cal Roberts 

Today’s big idea is: How will remote diagnostic tests change ophthalmology and vision care? It might be a foreign concept for some, but the specialists in today’s episode, Dr. Peter Pham and Dr. Sean Ianchulev, founders of (Keep Your Sight, a nonprofit focused on remote diagnostic vision tests) share how they can conduct more reliable perimetry tests that help detect macular degeneration, glaucoma, and other conditions that lead to vision loss and eventually blindness — remotely, while patients stay home. Developments like these in remote diagnostics are a stepping stone for the ways machine learning will impact the field of ophthalmology in the future. This episode also features Dr. Einar Stefansson and Dr. Arna Gudmundsdottir, developers of the app, Retina Risk, which helps with remote risk assessment of diabetic eye disease for people with diabetes, as well as Shirelle Jones, who lost her vision due to glaucoma.

Podcast Transcription

Jones:  I played softball.  I played basketball.  I did everything a teenager throughout her life could do.

Roberts:  Shirelle Jones started having vision problems as a teenager and was diagnosed with glaucoma in her mid-20s.  Glaucoma, a disease that causes peripheral vision loss, has no cure, but early interventions and diligent management can slow its progression.

Jones:  I worked as a collections representation, a senior collections representative with AT&T.  And I did that job for almost 22 years.

Roberts:  Like too many, though, Shirelle wasn’t able to keep up with the medications and appointments required to manage glaucoma.  She was in her late 40s when her sight worsened substantially.  At that time, she did everything she could think of.

Jones:  I went through every doctor.  My main doctor.  I even went all the way to Columbia, South Carolina at one point in time to see if they could help me with my eyes.  Every little tidbit that someone mentioned as far as “oh, they may be able to help you.”  And this is the God in Heaven’s truth, I went to nine eye doctors in one month to see if there was anything anyone could do for my eyes.

Every doctor I went to told me, Miss Jones, I’m sorry.  There’s nothing we can do for your eyes.

Roberts:  Shirelle’s story is so common.  It is a struggle for people working and raising families to keep up with a lifetime of regular doctor’s appointments, testing and prescriptions.

I’m Dr. Cal Roberts and this is On Tech & Vision.  Today we’re talking about Telehealth.  Telehealth and Telemedicine both involve the exchange of medical information from one site to another through electronic communication and allow the provision of health-related services remotely.

As doctors, we have historically believed that we cannot diagnose and treat our patients without examining them in person.  And the big idea we are discussing today is technology to allow eye patients to do parts of the clinical examination at home and electronically send the results to their doctors.

Other areas of medicine have been able to respond faster to this trend than ophthalmology.  Today patients can take their blood pressure, their pulse, their blood oxygen levels, their blood glucose.  And now even an EKG at home and send it to their doctors.  But in eye care to date, the necessary testing has always been required to be performed in doctors offices.

Now telehealth is opening the doors through better self-management and better diagnostic tools so patients can work remotely with their doctors to prevent vision impairment, for example from diabetic retinopathy.  Einar Stefansson and Arna Gudmundsdottir are the Icelandic doctors and developers who founded RetinaRisk.  RetinaRisk is an app that allows patients with diabetes to manage and evaluate their risk of developing diabetic retinopathy.

Stefansson:  The initial idea was to individualize the approach to diabetic eye disease and then, from that basic thesis several avenues have grown.

Gudmundsdottir:  We started back in 2009.  We were a little bit maybe a head of our time in two ways.  Number one, people were not used to this idea of using algorithms a decade ago.  We were sort of new in the medical field.  It was not very, to say, an obvious idea for physicians.

Roberts:  The algorithm they developed takes into account the risk factors for diabetic retinopathy that have been published in the epidemiological studies.

Stefansson:  It’s very simple.  How long have you had diabetes?  Which type, Type 1 or Type 2?  What is your blood glucose?  What is your blood pressure like?  Gender has a role and whether you already have retinopathy.

Gudmundsdottir:  Some of these factors are not changeable.  You cannot affect your duration of diabetes and which type.  But the other factors like the blood pressure and the blood sugar control you can actually change.  It depends on how you manage your disease.  And that’s also the beauty of it as a teaching tool.  To be able to show the patient, well, if the diabetes is very well controlled you’re not going to go blind from this disease.  If it’s poorly controlled for many, many years, that significantly increases your risk of having sight problems.

Roberts:  In eye care, testing has always been required to be performed at the doctors’ offices.  My guests today are trying to change that dynamic with KeepYourSight.org, a nonprofit foundation that has created the first telemedicine platform for population eye health.  Dr. Peter Pham, an Ophthalmologist from Houston, Texas and Dr. Sean Ianchulev, and Ophthalmologist from Mt. Sinai School of Medicine, developed online tools for testing of peripheral and central visual fields to detect vision-threatening diseases such as glaucoma, macular degeneration and diabetic retinopathy.

Welcome, guys!

Pham/Ianchulev?  Hi. Great to be here.

Roberts:  Thank you.  Before we get into the technology they’ve developed, I asked Sean about his first name, which I’ve seen written as Tsontcho.

Ianchulev:  So Tsontcho is my first name.  Tsontcho Ianchulev.  But Tsontcho was very difficult, and the moment I started adopting Sean then suddenly people always stopped me and said, how do you spell that?  Sean or Shawn or Shaun?  Oh no!  That was not a good choice.  So I said, okay, I’ll stick with Sean.

Pham:  Actually, Sean and I have very similar problems with our original name because my original name really wasn’t Peter.  It was actually An, and the An in Vietnamese is actually spelled An.  Of course, when you tell people you’re An, everyone says oh, you’re a female – Ann.  So, at the end of high school I decided my name is going to be changed to Peter, which is also my Saint’s name.

Roberts:  Great!  So, these technologies that we’re talking about.  Let’s talk about them first from the patient’s experience.  In most cases when patients come to the doctor’s office, they put their chin into a machine and some lights go on or lights go off, or something is moving around.  At the end of it results are achieved.  How is this experience the same in some ways or different?

Pham:  When we first started with telehealth we wanted to basically emulate what is currently available without recreating a whole new landscape where patients wouldn’t be familiar with what they’ve done in the past in the clinic.

Ianchulev:  Think about something that we’re all very familiar – perimetry.

Roberts:  Perimetry measures a person’s field of vision.  Peristat is KeepYourSight’s online perimetry test for visual field losses caused by conditions such as glaucoma.  It’s a test that patients can actually complete at home.

Ianchulev:  What does it take to do a perimetry test today?  The patient has to come, to wait, to register.  They have to be taken by a technician to a dark room.  Positioned, instructed, and then after 20-30 minutes in the office, in the office chair you get a result.  And 50% of the time, especially on a new patient, that result is unreliable.  Now they have to do it again.  What a waste.  Why can’t they do it in home when they’re relaxed.  When they can repeat it.  And then the data and the results are accessible online.

Pham:  The hardest thing for visual field is actually getting an appropriate distance away from where you’d like them to be.  We know that our blind spot is actually about 15 degrees away from fixation.  With this simple trigonometry you can now use that blind spot to help position the patients correctly in front of the computer monitor.  That’s the first kind of advancement we did for achieving the goal of using online technology to perform a visual field test.

Roberts:  Explain just for a second what you mean by blind spot.

Pham:  The blind spot basically is the area where we can’t see and the brain kind of fills in the gap.  Let’s say you put a flashing dot on the computer monitor and you have another dot that’s in the center of the computer monitor.  When you look far away you can see both dots.  But, as you move closer, and you reach that approximately 15 degrees away from the blind spot, or the flashing dot onto the side there, it actually falls into that blind spot and one of the dots would disappear.  That’s how we’re able to position patients accurately in front of the computer monitor.

Roberts:  So, then we have the patient positioned in front of the computer monitor.  They’re looking at their screen.  Now what happens.

Pham:  We basically provide patients with a random sequence of stimulus.  Depending upon their response to the light stimulus, the algorithm in the test would advance to the next stage and to the next stage and the next.  Similar to what is done with Humphrey visual field which is the gold standard in eye care.

Roberts:  Can we do this with any computer or do we have to have a special computer?  What’s the requirement?

Pham:  It does require a monitor that is 12 inches or above in order to not use quadrants of that monitor there.  With that being said, the monitors are getting bigger and bigger and of higher quality every day.

Roberts:  This is a test that is used often by patients with glaucoma, is that right?

Pham:  Right.  And not just glaucoma.  Basically, almost any disease that affects both the peripheral space and also even for macular disease.  We now have developed a system called Macustat which is testing the central visual field space which Sean can talk more about as well.

Ianchulev:  As Peter mentioned Macustat is another great technology because it does a complex, multiplex testing of the macular function which combines visual acuity, dynamic grid perimetry, photoreceptor stress testing and hyper-acuity perimetry, with scotoma registration and metamorphopsia registration for progressive monitoring. That’s another huge that would really capture huge utility with patients for dry, wet AMD, diabetic retinopathy, diabetic macular edema.

The goal here is to build a platform, just like Netflix for movies, where we can have many validated clinical applications, Software as a Service.  Whenever possible, I think they always stay along the lines of digital technology, clinical application that is hardware free when you merge that and put that on the backbone of the telemedicine platform you can achieve exponential population help.

Roberts:  The patients take these test and the information then gets transmitted to their doctor.  What’s the doctor looking for when the doctor looks at these tests?

Pham:  The doctor would be doing the same thing he would do in his clinic when he is sent a visual field from a Humphrey device.  He would look for certain patterns and defects that would signify either a certain disease, whether that’s from a pituitary tumor, from glaucoma, or if he has an existing patient he wants to know whether that change is stable or whether it is worsening.

Ianchulev:  It was our goal to do a hardware-free, digital/virtual device.  We felt, in ophthalmology, we’re kind of lucky.  We don’t have to do a colonoscopy.  We’re looking at visual function.  So perimetry lends itself to a fully virtual Software as a Service device.  We can interrogate your visual field, map out your contrast and be able to make very good results, get much more reliable results than what you can get into the office.

Roberts:  This is a big advancement in telemedicine for eye health.  Imagine what at-home testing could have done to track and possibly delay the progression of Shirelle Jones’ glaucoma.

Jones:  There wasn’t all of what is available now for me back that.  Now you have technology that helps people with vision problems.

Roberts:  Both Peter and Sean have made massive contributions in the field of ophthalmology and eye care.  They have both been parts of new drugs, new devices.  They’ve started companies.  They’ve both had a major impact in the way that patients are cared for and all of us eye doctors are so appreciative for what they have both done.  I asked them, why did they choose tele screening?  Why is it important to them?

Pham:  This really started when we were both training at Doheny Eye.  This was back in the late 90s, early 2000s.  We were training at a county hospital in L.A., L.A. County.  Like any county system there, there was just an overwhelming amount of patients and need.  We were working in the clinic and what we’ve noticed is the fact that doing just routine visual fields would take literally weeks to months for patients to get in.  A lot of these patients would have fairly advanced glaucoma that really needed a good clinical assessment in order to help optimize the treatment plan.  There’s only so many visual field machines that can occupy an office.

Ianchulev:  It really came out of an unmet need that we saw as residents.  We were taking care of indigenous people.  We had patients going blind waiting for weeks, sometimes, to get perimetry.  So, when we looked into the opportunity here, we donated the technology and we decided to go that path many years ago when there was no reimbursable pathway for telehealth.  Maybe because we both came from Communist countries; both came as immigrants here.  We realized the importance of public health.

Roberts:  Sean, let me ask you about the potential impact of technology such as this.  I can think of two areas.  One, it will reduce the amount of time and work that has to be done in a doctor’s office.  The other, it brings technology to areas, locations, continents, populations that ordinarily wouldn’t have access to this level of technology.

Ianchulev:  It means that you can deliver a way to test something anywhere, any place, any patient, in home, in the office, in Tanzania, in Ethiopia, in Europe just like you do with many other things.  Like you do with some of our other software applications today.  I think telehealth has been relevant for many years.  We were a little bit ahead of our time when payers and everybody else were kind of resistant to that.

Because the concept of really extending the doctor influence and extending the doctor visit to the home is way before today’s age.  That’s how medicine started.  The doctors were actually paying a visit.  So, I don’ think we were groundbreaking in any other way, it’s just that I think society is catching up now.  Especially with COVID, we realize the big disconnect that exists between medicine and public health and the need to really get to the patients in many ways other than just taking a trip to the doctor’s office.  Because we’re using Software as a Service cloud technology, and that can be done with many tests.

Roberts:  One of the themes that we’ve been discussing in this podcast series is the role of artificial intelligence and programmed learning in the development of technology that improves the lives of people who are visually impaired.  It would seem to me that these tests that you’re doing are just primed for programmed learning to be able to interpret the results, and therefore you could be able to screen large numbers of patients and only bring to the doctor those that are definitely abnormal.

Ianchulev:  That’s true, and through the years because we’ve done now for 20 years we have tens of thousands of tests in a database, they’re primed for all kinds of learning.  But, we’re still on step on.  Presenting data is reliable to doctors is step one.  Let human intelligence actually look at that.  That is a huge unshackling of technology in testing.  And then, you’re right, we can always get down like everything else into machine learning and get even better, and look at things we never were able to see before.

Now we can have an algorithm look at a human retina and tell you whether that’s from a male or female with a 95% accuracy.  You know that if you’re a retina person, or retina specialist who has seen 100,000 retinas in their lifetime, they still can’t tell.  so, obviously when we give data and put machine learning on top of the current status of human intelligence we can find things we never could do that.  And all of our data would keep inside this cloud-based Software as a Service, and it will lend itself in a HIPAA-compliant way to be able to do a lot of intelligent learning that we’ve never been able to do with perimetry.

Roberts: Sean is being a little bit self-deprecating when he says that what he and Peter were doing with Peristat wasn’t revolutionary.  The pair developed the Peristat visual field test in 2006.  Fifteen years before conditions would become ripe for it.  But, he is right about one thing.  something was standing in the way of having this service widely available to doctors and their patients.

Ianchulev:  As you remember, until recently, Medicare and the payers did not allow doctors to get anything reimbursable, any testing, unless it’s done into the physician office.  I don’t know why that has been the case.  When doctoring started, actually going to patient homes, I think that technology will help us get to the next level.  Technology has been around for this.  It hasn’t been applied for this.

Roberts:  One thing that changed is that in December of 2020, the Centers for Medicare and Medicaid Services finalized a change to this restriction allowing doctors to be reimbursed for remote patient monitoring.

At Lighthouse Guild, the pandemic created this opportunity for us to push a lot of our examinations from in-person to telemedicine.  And the necessity of the pandemic really pushed the transformation.  How has the the pandemic promoted or inhibited what you folds are doing at KeepYourSight?

Ianchulev:  We had it here on ground zero at New York Eye and Ear Infirmary at Mt. Sinai.  When we first encountered this surge of cases, and being here in NEw York was a unique place in March/April of last year.  We were seeing the worst of it and it came so quickly.  One thing that happened that I didn’t suspect is that I was getting so many calls from doctors who knew about what KeepYourSight was doing back in the days and they wanted to get access to our system in order evaluate their patients.  Because, right now, they couldn’t have patients come.  The patients, they didn’t want to take the risk.

And that really stimulated Peter and I to really say, this is the time to open the system.  Offer it more widely.  Make it HIPAA compliant.  Invest in the whole platform and really add all of the tests that would really make this a useful system to all eye doctors.  And you will see when that gets rolled out in a big way in the coming months, the Foundation will offer that free of charge for all of the places where impacted by COVID where they have issues and patients cannot come.  Because we don’t want patients with wet AMD, with advanced glaucoma to forego care and actually go blind while they’re afraid sitting at home.

Roberts:  Both of you came to this country from other countries.  Is that a coincidence that you two have become great inventors and entrepreneurs?  Do you think that there’s something that inspired you coming to America?

Ianchulev:  I never thought of that, but Cal, you’ve got two immigrant eye doctors here.  Maybe that’s why we have such a lasting partnership.  We both came to this country with no money and no support.  For me, that was at the age of 18 when I came here literally after Communism.  I landed at JFK with $200 in my pocket.  The empty pocket is a very adventurous enterprise.  Once you do that – I talk about this with my kids – once you do that in your life, I think very few things are very daunting to you as a person.

Pham:  I, myself, came to this country back in 1979 at the age of 8 years old.  I still remember the first day when I came to this country.  We actually landed in Michigan in the depths of winter and of course, we flew in from Hong Kong where it was like 70 degrees.  Coming to America really brings out the best in immigrants.  You know where you came from, and you really want to make not only your life better, but hopefully everyone else’s life a little bit better as well.

Roberts:  When I speak to innovators one of the things I always learn is that innovation does not occur in a straight line.  You don’t go directly from point A to point Z.  There are detours along the way.  Were there trial and errors?  Were there roads you went down that didn’t work and you had to start again and go in a different direction?

Ianchulev:  Not only have a I learn that non-linear pathway is the way of innovation, it’s very concerning for logical doctors because we’ve been taught to be very irrational and that makes us uncomfortable when we don’t really know where we’re going to end up.  In addition to that I would say the one thing we’ve learned is that if you’re ahead of your time – we started this technology 20 years and it was at the time of AOL and people were still saying “would elderly people ever go to the Internet?”  These are the kind of questions we were facing then.  When you are ahead of your time you have to be a Klingon.  You have to hang in tight and cling on to the time because sometimes it takes for technology to catch up.  We’re very happy that we didn’t let that fall through the cracks and Peter and I persevered, and we offered it for nonprofit screening through the years because right now, based on our experience and what we have, KeepYourSight is in the best position to develop the telehealth platform of the future.

Roberts:  The results of this innovation, of what you do with KeepYourSight, what you do with Zocular and with other companies is going to protect the vision of huge groups of people who might otherwise lose their vision.  What keeps you motivated?

Pham:  My main motivation in both KeepYourSight and Zocular for Dry Eye is to help people with diseases that are hard to detect, hard to treat.  If we’re able to find a solution that meets the need there, I think that’s really what keeps me motivated every day.

Roberts:  So, Sean, same questions to you.  Of all the areas of medicine that you could have gone into, why did you go into vision science?

Ianchulev:  For me it was a very simple reason.  The greatest ophthalmologist that I know is actually my mother.  She is a retina specialist.  She introduced retina surgery in Bulgaria back in the day.  She was a professor of ophthalmology and one of the most renowned ophthalmologist in Bulgaria and has lectured around the world.  I started out early just seeing how grateful the patients were.  It was amazing to see medicine and technical expertise blend into the hands of a great woman.

I also, I keep laughing, but what I really liked every time I went to her office was the smell of coffee that the nurses made, which was the not the case in my father’s office for engineers.  So, I said, wow, this is so great.  Then the patients brought so many goodies because in Bulgaria, doctors were always paid the same in Communism as anybody else.  The patients were so grateful they would bring chocolates and all that.  So, it was just amazing to see how good it is to be a doctor.

Roberts:  People become doctors to help others.  Sean and Peter, with their KeepYourSight Foundation embody this.  With expanding technological capacity in the hands of patients, and expanding coverage accelerated by the pandemic, KeepYourSight technologies are finally poised to allow doctors to help more people, and people who live further away and in more isolated locations.

But, even as this massive shift in medicine is happening.  Dr. Aarna Gudmundsdottir has one eye on the future, as we all should.

Gudmundsdottir:  If you look five years down the line I’m sure they’re going to have more AI coming along, and that means trying to integrate the imaging of the retina of the algorithm as well.  that will be a huge step.

Roberts:  With the technology like KeepYour Sight help Shirelle Jones manage her glaucoma?

Jones:  I really believe it would have in a great way because I could just take my eyes, and like you say scan your eyes and go into your eyes and see what’s really going on.  I’m not used to all of that way back then.

Roberts:  It just so happens that Shirelle relocated to New York from Georgia.  There were a lot of reasons for her move, but one of them was to be closer to medical care.  She’s found other upsides to being here.

Jones:  My thing is, I have never been in an environment where I knew that there were so many people that were going through the same thing I was going through until I came to New York and interacted with people at the Lighthouse Guild.  It may come down the road, maybe I’ll wake up one more morning and I won’t have no eyesight.  But just the simple fact of knowing I didn’t wait, I didn’t wallow in my illness.  I didn’t sit back an not try to live my life because of the fact that, oh, I might one day go blind.  No, you keep going.  You keep living.  You keep striving.  You keep believing.  If nothing else, believe in you.

Roberts:  Did this episode spark ideas for you?  Let us know at podcasts@lighthouseguild.org.  And if you liked this episode, please subscribe, rate and review us on Apple Podcasts or wherever you get your podcasts.  I’m Dr. Cal Roberts.  On Tech & Vision is produced by Lighthouse Guild.  For more information visit www.lighthouseguild.org.  On Tech & Vision with Dr. Cal Roberts is produced at Lighthouse Guild by my colleagues, Jaine Schmidt and AnneMarie O’Hearn.  My thanks to Podfly for their production support.