This Week in Health Tech

Artificial Intelligence (AI) in Healthcare - Part 1

January 07, 2020 Vik Patel and Jimmy Kim Season 1 Episode 4
This Week in Health Tech
Artificial Intelligence (AI) in Healthcare - Part 1
Chapters
This Week in Health Tech
Artificial Intelligence (AI) in Healthcare - Part 1
Jan 07, 2020 Season 1 Episode 4
Vik Patel and Jimmy Kim

In this part 1 of AI discussion on This Week in Health Tech, Jimmy and Vik comment on AI trends in Healthcare. Discussion of AI influence on wait times, health organization efficiency, and more importantly improving patient outcomes. 
We don't know how we ended up with a comparison of AI in StartCraft game and healthcare, but yes that happened in this episode. 

Support the show (http://www.thisweekinhealthtech.com/)

Show Notes Transcript

In this part 1 of AI discussion on This Week in Health Tech, Jimmy and Vik comment on AI trends in Healthcare. Discussion of AI influence on wait times, health organization efficiency, and more importantly improving patient outcomes. 
We don't know how we ended up with a comparison of AI in StartCraft game and healthcare, but yes that happened in this episode. 

Support the show (http://www.thisweekinhealthtech.com/)

spk_1:   0:02
Welcome to another episode of this weekend, HealthTech, where we cover the latest news and trends in the health tech industry. My name is Jimmy Kim, and I'm accompanied by my co host, Vic Patel. Vick. When you say hello, everybody. Hey, Jimmy, Welcome back. Welcome back. Welcome back. Welcome back. And Happy New Year to you, man. Happy New Year. Happy 2020.

spk_0:   0:23
Yeah. You know what? Today's topic actually is perfect for the year 2020. But let's talk about your holidays. How were your holidays?

spk_1:   0:30
All the days were really good. Um, as you know, I took I took some time off action. Took three weeks off of pretty much doing nothing. And, uh, I'm your This is This is my first day back now, have Where? Thanks, Vic. Thanks. Thanks a lot. Dog kidding. Yeah. So look, you know certain. I'm starting 20 Tony off with with a bang, and I'm ready to go. Ready to get this going? How was how was your holiday?

spk_0:   0:55
I hear that in your voice, By the way A little bit, You know, I hear the bang.

spk_1:   0:59
You hear the back? Actually, no, no, no. Tell Della

spk_0:   1:02
listeners a little bit quick. I know. You just told me what happened during the holiday.

spk_1:   1:08
I know. I know. For you. I know you wanted to be like I was out drinking and I was out partying, and I I wish I was. I wish I was. But I, uh I spent the last couple of days of 2019 at a water park with the kids Indoor Water park. And, uh, I unfortunately, I was the only one to walk away with a little Ah, little special guest. And I've been out of commission for last couple of days. And, uh, that's why. Maybe it's a little nasal, your little horse. So, um, yeah, I feel good, though. You know

spk_0:   1:35
what? You look good. You feel good. I mean, you sound better. Yeah. You're almost that. I would say you're 95% of the way, so yeah, our listeners will Barrett your voice this time. It's okay. You know you'll be You'll be much better next

spk_1:   1:49
time. You know, it just means that you can take the reins and you you can You can talk a little bit more than this podcast. So before we get into today's episode, I just want to give a special shout out to a very special sponsor. Tito Inc. Tito Wink is a trusted tech partner for health care organizations. For more information on how to partner With Tito Wink, please visit Tito Inc dot com. That's t I d o i n si dot com, and we actually got another sponsor. Sponsor is host my podcast dot com, where professional podcasts are made. Can you believe that? We had to sponsors?

spk_0:   2:26
Oh, it's all we got. Three.

spk_1:   2:30
We're moving up, moving up in this world of a moving up on this world And you know

spk_0:   2:33
what? I think we should add a new podcast sponsor with every new episode. So this is what our fourth So actually we are lagging behind Jimmy. It's Sunday night, and it's time for this week in health tech.

spk_1:   2:49
You know, it's it's it's This is unusual. Time for us is not the usual time that that you and I meet up. So, um, I'm expecting a little bit different flavor and in terms of different flavor, I think the topic that we are going to be talking about today is you know, God, I got a lot of spice to it got a lot of flair to it, which is we're gonna be talking about the impact of artificial intelligence in the health tech industry.

spk_0:   3:12
We'll talk about artificial intelligence in a very general in a general way, and talk about the trends that I see in health care and where I think we may go in the near future. And I didn't know what I mean from your point of view. Being an outsider in health care, What were you planning on or what is your focus for this episode?

spk_1:   3:36
Well, my focus for this episode will be honest with you, is to learn is to learn and to kind of get an idea of what's out there because I think what is in my mind, I'm a little scared to say what what what I'm thinking, but is prostate is probably just totally like way off. What? What actually is going on in the health chickens? You mean in my mind, of course, it's like it's like sky net. It's like terminators. It's ah, you know, walk

spk_0:   4:01
into the hospital and they scan your body and they're like, What's going on? You

spk_1:   4:06
know, very like scifi tech thing. I mean, everything that I know about artificial intelligence. To be very honest with you, this is through gaming. Um, I I'm a big I love watching games online. Like, uh, I'm a big star craft watcher Starcraft to buy local watching it and, uh, the most recently there has been a n a I who's learning how to play Starcraft. And I'm just learning about that, eh? I and how it's learning how to play, play the game and to beat, you know, top 10 other other players of this of this strategy game and one of the things that I that I picked up from it was that in a short period of time, this Aye, aye. Who's basically just learning how to play a video game has learned, has over has clocked in over 200 years worth off of playtime, has 200 years worth of knowledge on how to play this game in in a just a span of years. So I'm

spk_0:   5:12
joking. Doesn't go. Is it going through previous games and just analyzing the moves and just learning

spk_1:   5:17
it's going through previously? It's going through previous games studying games. It is it's playing against itself and working out different, different strategies. One of the things that I find very interesting when I watched these when I watched the eh I play the game is it does so with the 200 years worth of experience. It does all these things that it it it does. All these does all these things that are generally considered mistakes and and yet it yet it wins. Yet it beats it. Has most recently has beaten, beaten a top 10 player. Um, it is what,

spk_0:   5:57
That player from China?

spk_1:   5:59
No, I believe that player was from Korea, Corey. Okay. Yearnings to dominate the Starcraft world at the moment. And, uh, yeah, I just you know, I was talking about this. I'm thinking, OK, video games, you know, a big market, but probably nowhere near as big as the health tech industry. Right? So if they're doing that in the video game world, I can on Lee, imagine what they're doing, what they're doing in the health tech world.

spk_0:   6:30
It's Yeah. I mean, you were seeing healthcare being so big should have a lot more implementations off the I, but again, healthcare, it does move a little bit slower right than every other industry. And and we have talked about this in our past episodes. Um as many comes to technology and using the latest technologies. Healthcare is generally behind than other industries. However, in this case, I think I is going to have a huge impact, and there's more and more investment going on in the ice. So today, Jimmy, there's about six billion in investments for a I Okay, and what they're predicting is, by 2026 it's going to increase $250 billion investments in the eye. I mean, that's that's a huge So you can imagine. I mean, people are moving towards Maur and more applications in health care, and the biggest goal is obviously improving patient outcomes. And so let me give you this statistic and see what your reaction is. Medical Enters is the third leading cause of debts in America, third leading cause. So about about 10% 10% of all deaths are because off medical errors.

spk_1:   8:05
And when you talk like medical errors like what kind of errors are we talking about? Like the doctor forgot, you know, that he left the scalpel in somebody like those errors or they extreme that maybe miss read something on the charts or or something like that. We're talking like those kinds of errors.

spk_0:   8:24
Yes, I think the second example that you provided mis read the chart. That probably happens a lot. And the other thing is about, let's say medications, right? So if you are prescribing different medications and you did not catch that Hey, these two medications that I just gave to the patient, they actually do not interact well together, so it may lead to something else. And the other thing is, if the diagnosis is not done correctly So you actually treating the patient for A And actually the patient has a disease Be, well, the whole time that you raised it, treating the disease that you thought the patient had, they actually didn't have it. You know that a condition that got worse. The diagnosis is definitely, you know, that's what I think. He I's gonna have a huge impact diagnosing the patient correctly.

spk_1:   9:23
And then how do you like maybe, maybe, like, walk me through? Like how? How could a I catch those mistakes like, is it you know, Is there some sort of software you think that's gonna be used? Or it's gonna be through our smartphones or what? What do you think? Definitely a

spk_0:   9:41
software, and definitely similar to the example that you gave off the eye in the game. In this case, it does something similar. For example, I was I was checking out a couple different symptom checkers just to see how good they are in terms of OK, if these are the different symptoms that I have, how good can you diagnosed my condition? There's buoy held in Babylon held systems. You know, a couple of them. I think what they do in this case is they actually have a chat. Bott, that listens to your symptoms. So it's It's when I say, Listen, you actually are providing them your symptoms. I mean, you know, you type your symptoms to a chat, but obviously and and it asked you follow up questions, I had to answer quite a few questions, because what? It does it as soon as you answer one question, then it comes up it. Okay? These are my five follow up questions. Provide me this information. Okay, Now we provided that Okay, I see what's going on. You have fever. Last couple days, you have been coughing, but it's dry. Cough goes into the next set of questions, and one of them would be, um how are you sleeping? Are you able to sleep? Are you able to breathe? Okay. You know, so just just goes on with a bunch of questions until it finally tells you. Hey, you may have one of these three conditions.

spk_1:   11:15
So you think this is just like the dyslexic help MD deductible last Google? Exactly. It

spk_0:   11:23
seemed a little bit like refined Google, right? You are actually giving them the information and this gives, you know, instead of searching your kind of providing information, and then it's using its knowledge base in the background, You know, it probably has a huge data source. It's coming up with all these conditions, But one thing kind of stuck out to me. You're just funny, because at the very end, one of them asked me, What kind of insurance do I have in what location? You know what we're doing live? Yeah. And I'm like, Okay, this goes back to our episode two off security there, You know, we talked about using different data points even without having the specific name or address information you can actually say. Oh, this, You know, if Jimmy comes in the office

spk_1:   12:22
tomorrow Whoa. Don't use my name. Use you use your name. But if they isa listening, I want him to come find you. Cannot be that

spk_0:   12:29
everyone knows our name. So anyway, you know, that was that was a little funny. It's nothing spectacular. It's not like you're standing in front of a camera. Maybe this may happen in future where you may actually stripped down and stand in front of a camera. And it's analyzing your body, which actually, I think instead of that, we may have a e. I think we talked about this. We may actually have a chip in our body that's constantly checking and monitoring your body to CEO. What are the symptoms If I'm seeing anything if normal than all I'll note that And if I see more and more abnormal, you know, if I see certain data points that are not actually registering as normal, then maybe I'll notify the physician.

spk_1:   13:17
Yeah, and you know, kind of going with that. Actually, actually did read an article where there is a There are research researchers in the United Kingdom, and they have kind of they're there in the works of developing a tool that can identify development of diseases by analyzing images of of Children's faces. I actually read that not not too long ago only unlearn related. It just kind of popped up because maybe I was looking up something for my, for my kid kid's symptoms and basically, whatever Go

spk_0:   13:51
going held,

spk_1:   13:52
I was googling health. Yes, I was googling health, and it came up and basically what it was and I mean makes sense that we do to do this already, but taking selfies or taking pictures of our skin or of her face and then and then it being able to kind of diagnosis just in terms of dermatology, purposes of what? What you may have.

spk_0:   14:18
And you don't want Jimmy. So you mend me started and you kind of indicated you into the water park. And you caught some kind of bug right? Now think about this. You didn't go to the doctor. You're a tough guy. And you know, you just laid in bad, and you're you're okay. Now,

spk_1:   14:36
if you ask my wife. I was definitely not a tough guy. Lost you? You know, that was just for

spk_0:   14:42
the listen, I know you're not

spk_1:   14:44
tough way are working out small Ravix So

spk_0:   14:51
going back to getting back on the topic. I mean, let's say you didn't feel like going to the doctor, because you you think OK, yes, I caught something, but I'm just going to rest. However, wouldn't it be nice that if you enter some kind of symptom checker that we are talking about and maybe using a camera, Or maybe, you know, answering bunch of questions and it would actually tell you? Hey, this is what you have and you can actually go and get this medicine over the counter. That may help you. This is an example, right? Kind of those. If you need a medicine that's not over the counter, then obviously, you know. Okay, maybe I should go to the doctor and get this, and it will make me feel better. So that's the other thing off where a I can help. When? What's your biggest? So why didn't you go to the doctor, is it? Tell me honestly.

spk_1:   15:47
Well, I'll tell you for me, it's It's, I guess, one. I didn't think it was very serious enough for me to go to the hospital, which kind of leads into number two, which is I'm already feeling like, super crappy. The have to sit and wait in the e r. Or have to sit and wait for my doctor's appointment and to even get there. I was just like, just not having it like I didn't want having any part of it just for I'm gonna guess, just for the doctors say, Oh, yeah, you got a cold. You got to go and you gotta You gotta go home and sleep because that mean that's happened to me. I'm sure that's happened to you a few times when you like him, and I hope they give me something that's gonna make me feel better. And

spk_0:   16:28
they seem disappointed. Oh, yeah, I wish there was something bigger. I just waited in the waiting room for two hour.

spk_1:   16:36
It's the worst one, like there's like all these other people that go in before you, and then they're in for, like, 30 minutes, 40 minutes, and then you go in and the car I wanna get a nice, like 30 40 minute consultation Made me feel better is like they give you, like, open your mouth. I'll take years, chicken guys. All right, go home and rest. You're good. You're just like, Oh, man,

spk_0:   16:54
drink lots of water, by the way.

spk_1:   16:55
Yeah, yeah, yeah.

spk_0:   16:57
And so that's it. The word, the w word, the waiting, right. I mean, that's That's one of the things that I think e I should definitely help because what's gonna happen is based on who's waiting in the waiting room. And, you know, based on your symptoms that it already has collected, maybe you collect the symptoms before you. You get there either it through an online check in or once you get there, maybe you enter that information at a kiosk. However, you get that information now, it knows these are the different people waiting, and I can actually prioritize them based on their condition. You know, obviously, if you're talking about a yacht in the hospital, you are going to take the person who is was a stab wound instead of, you know, and it may be dying instead of someone who is just having a fever

spk_1:   17:54
you got sick at a water park, right?

spk_0:   17:56
And got sick at a water park. So the wait times is definitely one place where e I can probably see how many people are waiting. Maybe what it can also do is make calls to additional employees. Right? So if you're in the hospital and you are seeing okay and the other thing too, it can if if the ambulances are plugged in and we know Okay, there's another five ambulances on the way. And based on the 30 people that are already waiting, you know what we need additional Dr Marty and so many nurses. So let's make those calls. Let's get additional staff in and help these patients. So again, you know, it comes two improving the patient outcomes at Leanna. So yeah, I mean all the way from, you know, the symptom checker or finding, you know, diagnosing the patient correctly to reducing the weight times. I mean, those are two amazing examples,

spk_1:   18:58
you know, and kind of kind of. I mean, I mean, I hate to kind of bring it back to the video game thing, but it's it's it's really like all I really know about a I but one of the things that that a I was doing in that Starcraft game was which everybody found extremely unusual was in the game. You have to create these workers that have to go that have mine, you know, mine mine minerals for took to produce Ah, worker, mother, other soldiers and things like that. But one of the things that I was consistently doing was it was producing an immense amount of workers. Immense amount of immense amount of workers to mind the mineral fields. It didn't make any sense until until until one of the main strategies was for for the other enemy, for the for the other person to go in and to consistently attack the workers. So then the work account would be down, and then your mineral camp will be down and so on, so forth and you can build less workers. But the A. I had produced so many more workers that even though it kept getting hit from right and hit from from the opposing party, it was still producing a high amount of minerals to be able to produce more and more workers. So it took a

spk_0:   20:08
very different approach than people Who are all these champions who have inventing the game. And maybe because it looked at mountains and you know all this data

spk_1:   20:22
200 years worth

spk_0:   20:24
200 years worth? What? What do you mean? How is it? 200 years words? Panel. You've been combined. Combined. Yeah. Okay, okay. I'm like, this game's been around for tuna data worth 200 years, and it came over the strategy. So maybe the same thing can be done here. There, you know, let's say cancer. I mean, we have been dealing with so many different types of cancer, and and it's gotten a lot better, you know, before, when you said the cancer were you pretty much, you know, we're preparing. I mean, this is and it's not even that long ago, right? 2030 years ago. But now, cancer, lots of cancers are very treatable. However, there's still a lot more that we still don't have a way to treat them. And maybe using a I maybe once it looks at 1000 years waters off data, it can come of it. Oh, by the way, why? Why don't we try this, right? I mean, don't be an amazing improvement and patient

spk_1:   21:29
up, basically catching things that we have probably looked at and stared at, You know, all these times and or probably just have thought, No, that's not gonna work. And

spk_0:   21:41
maybe we haven't started because if he, you know, I mean, we have a lot of doctors coming out of schools every year all over the world, and there's a lot of research going on. But think about this. Let's say you have 10,000 researchers working on finding the next cure for cancer now with the eye and the resource is that we can give it with the eye. You're pretty much multiplying those researchers by 100 right? So instead of only 10,000 what is it now? You got a 1,000,000 people researching,

spk_1:   22:24
okay? I mean, for me like this, this this all sounds nice and rosy and like, it's amazing. It's gonna do all these amazing things. But again, like if I If I could be one of the two of us, you know, that could perhaps play devil's advocate or or So, um, how reliable is this going to be? How how much faith should we be able to put into a I and again kind of maybe put on my put on my aluminum tin foil hat here. Ah, should we be putting this much faith into artificial intelligence?

spk_0:   23:02
Well, you'll be excited with this, that eggs enger things that most of the investment in A I is going to go into robot assisted surgery. That's okay. Talking about putting in trust. Yeah. I mean, you know, doctors using robots to operate on you. You know what? Maybe it's a good thing. It's definitely a good thing, right? Instead, off a single doctor making decisions on your body. And now you have this robot that has all this knowledge Based on what it's learned from thousands and thousands of previous surgeries, It's actually helping the doctor as they're working on you saying? Yep. You know, that's the right move. Go ahead or stop. That's the wrong move. Don't Don't do it. Yeah,

spk_1:   24:02
but then that what a boat. You know, what about that? Like human like intuition like that, You know that. You know, you know that gut feeling that

spk_0:   24:10
you've been watching too many.

spk_1:   24:13
Okay, Maybe, Maybe, very. You know what

spk_0:   24:19
I mean? Okay, so you will have that I mean, that's so what I just explained to you about this surgery, right? I mean, this example off the robot, assisting the physician in a surgery they want. If you were, let's say you were getting Let's come up with something really bad.

spk_1:   24:41
Yea, no. Okay, let's look at the example that let's say like something Harding happens. Something happens. And it's like, All right, Sorry. Yeah, Jimmy, Jimmy's heart failed. And, ah, through all my thousands of hours of thousands of years worth of research, there's no chance to resuscitate him. Case closed. Boom. Right. Whereas I'm hoping maybe a human being will be like Okay, you know what? Let's try. At least let's try something, you know,

spk_0:   25:12
You think Hey, I was just gonna look at your data points and say Nope. There's not a lot of hope.

spk_1:   25:18
Yeah, I

spk_0:   25:19
asked on this person. Look, you know what? That's actually that's a good point

spk_1:   25:22
in my life. If I wasn't a I'd be like, all right, you know what should I spend the extra two minutes and energy and resource is on? This guy that I know is is gonna He's done. Should I move on to the next and spend that two minutes and try to save this next person's life. Saved my life, man.

spk_0:   25:40
Yeah, you got me there. I mean, that's a great That's a great example of yes, no emotions involved, right? I mean, a I's not gonna Maybe maybe in the future and there really advanced. I mean, that's what separates us and robots, but that's I don't know. I mean, yeah, I think for now we'll just say yes. Jimmy's done. I don't think the heart transplants gonna help him much. That's more one. Let's move yet you later. Later. Sorry, you're done. So I think that's that's why What is not working by themselves? You know we are. It's actually the other way, e eyes assisting us. We're needed. We're not at the point where doctors are actually going to maybe watch and and assist very needed. I think it's still the other way, you know, it's still we're going. We're going to use a I where necessary, and Larry can help us. But there's no way we are. There is actually making 100% of the decisions.

spk_1:   26:45
Well, I mean, I I hope I hope we come to some sort of medium soon. But I'll have to say, I'm I see a lot of good with it, though, you know, despite despite my whole 10 4 hat, hat theories and stuff like that, I mean, I can only imagine, like, one. Uh, I'm hoping maybe you could tell me otherwise. Hopefully you can't tell me otherwise, but perhaps with the eye coming in, it makes a lot of these. A lot of service is perhaps lost a lot more cost effective that a lot more people can can get seen or can get treated. You know, going back to be going back to patient care is the number one thing, right?

spk_0:   27:24
Yes. So I think it will definitely increase the efficiency in organizations. And we talked about the wait times, right? So and the other way to like, if you if you are in a clinic and looking at the schedule and looking out and what's going on and how many people have called in, you cancel the appointment this and that. Maybe it'll also adjusted staffing accordingly, so it should help organizations save money by scheduling staff according to the calendar. Right, according to what's up coming And what the situation is at the clinic. One of the hospital, but the other way, too. I mean, you just gonna mention about okay, As a patient, hopefully it helps to reduce my cost. My hospital visit, you know, if I was admitted in the hospital, So Google Deepmind actually plugs in and I don't have the exact statistics on very Google. Deepmind is being used today. It's probably being researched at some of the big research university hospitals, But Google deepmind worse with the HR, the electronic health record. And when you were admitted in the hospital, it scans your chart. It looks at all the tests that have been ordered, the results that have been that the results off those tests and it makes a prediction within 24 hours off your situation. So it's going to diagnose and come off it. Okay, this is what's going on with this patient, and it's going to predict whether you are going to stay in the hospital longer or if you will be discharged within the next five hours. So that actually helps the hospital from not just staffing, but also, you know, just just seeing what actually Google Deepmind with all its knowledge base and all the data that it has collected from all the previous admissions in the hospitals, it will help the doctors in a way of Okay, let me compare from what we are thinking to what Google deepmind, as came, has come up with. And hopefully so that should also increase the efficiency.

spk_1:   29:47
And what's crazy too? You know, you just mentioned Google deepmind like I don't know much about Google deepmind, but Google Deepmind is the same. I think the same company that is in charge of that this video game, eh, I that I was telling you about

spk_0:   30:01
Oh, there you go.

spk_1:   30:03
It's crazy. They're just digesting it a little

spk_0:   30:06
in its part time when it's not looking at the patient charge, it's playing games. Just

spk_1:   30:11
playing came. I'm hoping. I'm hoping it's a different team, though I

spk_0:   30:19
predict that this patient hold on, I need Thio build someone working.

spk_1:   30:24
Okay, Well, look, I you know, you and I we're just We're just discussing this with this, I think just just the tip of the iceberg. I know that we are working on guests and experts to come onto the show who will who will probably who will shed, you know, more light on this and to educate the both of us on this and for me. I know.

spk_0:   30:47
And they have actually are using a I in some other applications or are planning on using a I and have firsthand experience. I mean, I have I have first hand experience dealing with products that are using a I because we do integrated lot of applications, for example, Ah, one that would actually see all the medications. So one that actually monitors all the medications that are being prescribed to the patient just to make sure that is no negative interactions that would actually not you know, that would not help the patient, but yeah, you know, we we will bring in Ah, for our part, too. Off the eye or all right, we should call it E i Episode two. Well, come over the better name.

spk_1:   31:36
I hope so. Well, that was one. I was fun. You know, I I'm glad. I'm glad we get to start with this. You know, the are very first episode of the new year that we're talking about things that are happening now and things that will affect us in the future. It's It's, I think, for not just this weekend. HealthTech but for for everything. I think this is gonna be a fantastic year. And, uh and I wish you and I wish all listeners out there a very happy New Year. Hope you make this your best new year and you, haven't you? How do you have any closing words for us? Where's the wisdom?

spk_0:   32:16
No, you said it, Jimmy. Happy New Year to all our listeners and I predict using Google mind that you are going to stay up. Stay out of the hospital this whole year.

spk_1:   32:28
I hope so too, brother. Thank you so much. And listen, make sure that you hit that subscribe button so you don't miss a beat. Hit us up in the comments field. If you want us to touch on the subject, if you want us to bring in an expert, we will do our very, very best. But please, please let us know in the in the comments an apple, an apple in in Spotify and Google wherever. Wherever you listen to this thing, er hit us that we would love to hear from you so until next time. This is Jimmy Kim. And

spk_0:   32:59
this is big hotel.

spk_1:   33:00
Good night. Even that in there. Okay. Good night. Here. What? Yeah, it's what?