November 27, 2020: It's basically human versus virus right now. Unfortunately healthcare has been paralyzed with a lack of data and information. Dr. Anthony Chang, Chief Intelligence and Innovation Officer for CHOC and Founder of AIMed is defining AI enabled solutions to ensure the healthcare sector is not left behind. What different types of AI are we seeing in healthcare? Where in the world are we seeing the most innovation? Has the pandemic impacted AI’s progress? What data and technology will we have in 20 years from now?
AI During the Pandemic with AIMed Founder Dr. Anthony Chang
Episode 334: Transcript - November 27, 2020
This transcription is provided by artificial intelligence. We believe in technology but understand that even the smartest robots can sometimes get speech recognition wrong.
[00:00:00] Bill Russell: [00:00:00] Welcome to this week in health IT where we amplify great thinking to propel healthcare forward. Today we have Dr. Anthony Chang with us to talk all things AI. Dr. Chang is the founder of AIMed. He's also the Chief Intelligence and innovation officer for CHOC children's in Southern California. So we're going to give AI a grade during the pandemic. We're going to take a look at where AI is going to go, moving into next year. So looking forward to that, my [00:00:30] name is bill Russell, former healthcare CIO, CIO, coach consultant, and creator of this week in health IT. A set of podcasts, videos, and collaboration events dedicated to developing the next generation of health leaders.
[00:00:41] I want to thank Sirius Healthcare for supporting our mission. And the mission of the show to develop those next generation of health leaders. their weekly support has given us the ability to expand and to grow our services to the community. And for that, we are extremely thankful. one last thing before we get to the show, and that is we've opened up our sponsorships [00:01:00] for 2021. So if you're listening to this show saying, Man. I wish I wish our company. We're getting our message out on this week in health. It, this is your opportunity. We've opened up those sponsorships for next year. If you want to, get a copy of our offerings, go ahead and send me a note firstname.lastname@example.org.
[00:01:19] I will respond back, give you our package and you can take a look at it. We would love to partner with you to get your message out to the community.
[00:01:26] All right now, Onto this discussion [00:01:30] about AI. All right. So today we are going to talk about AI during the pandemic, AI in startups, AI in use wherever else we happen to stumble upon today's guest is Dr. Anthony Chang, founder of AI med, chief intelligence and innovation officer for CHOC. Actually, there's a longer title there, but, well, welcome back to the show Anthony,
[00:01:53] Anthony Chang: [00:01:53] I remember with fond memories, our chat in the in Newport beach,
[00:02:00] [00:02:00] Bill Russell: [00:02:00] That, that show's been downloaded, quite a number of times, well over 1500 times. So people have, I've gotten a lot out of that and, and, I appreciate you coming on. So chief intelligence and innovation officer for CHOC but it's more than that, right?
[00:02:16] Anthony Chang: [00:02:16] Well, the founder and director of the medical intelligence and innovation Institute or MI3 for short. And it was from a generous grant [00:02:30] from the Michelle LUNG foundation or the Sharon Disney lung foundation that is supportive of the artificial intelligence agenda for not just children, but for adults and medical care in general.
[00:02:46] Bill Russell: [00:02:46] Wow. yeah. So you, you've really immersed yourself in this AI space. You started AI med, you're still a practicing physician. let's start with the [00:03:00] AI med ha how's the work in AI med going?
[00:03:03] Anthony Chang: [00:03:03] We had to make a big pivot because we were, as you recall, live events and, but we had already had pivoted to some virtual events and webinars and, network formations for people interested in a certain sub area under AI and medicine. And, and my book finally came out. I think we were talking about it last year. So far, I think I'm well received because I think [00:03:30] clinicians really didn't have a textbook that spoke to them in this space. Some of the other books are more geared for data scientists and AI experts, but I really wanted to create something that was made more for the non data scientists, and non AI experts. So, I think that's going to really, create a sense of urgency as well as, some validation for this area.
[00:04:00] [00:04:00] Bill Russell: [00:04:00] Yeah. So what was your, what was the core of the book? What's your message to physicians at this point?
[00:04:09] Anthony Chang: [00:04:09] That it's a very robust resource that we should take advantage of and healthcare, as other sectors in society have done. So, and, but with the clinician being a partner, rather than being delegated to, in terms of using AI tools in healthcare,
[00:04:28] Bill Russell: [00:04:28] Yeah. And that's what we talked about [00:04:30] before is AI as a partner. It's not the AI is not the clinician, but it's a partner to go and gather up the information process, that information, present some, potential, I dunno, points along the way to help. The clinician to maybe operate a little faster to process a little bit more information to handle the burden of the daily interaction with computers. I [00:05:00] mean, that's what I remember I was talking about is, are those primarily what you talked about in the book?
[00:05:05] Anthony Chang: [00:05:05] Yes. That's the consistent take home message is, let's see use, data signs and what it's capable of doing to the advantage that it offers and in working in partnership and in synergy with clinicians, not against. So, I think what's been nice is I'm no longer seeing a lot of papers that [00:05:30] compare computers with clinicians. I always thought that was not helpful. And what I'm seeing more and more now is clinicians with data science versus computers without, clinicians without that help.
[00:05:44] And I think it's, somewhat analogous to you driving. You can easily drive without a GPS, especially if it's somewhere close and you're very familiar, but most patients are complicated and, it does help a great deal [00:06:00] to have the support of data science. It's like driving with a GPS to a place that's far, and that you don't know, you're not familiar with,
[00:06:10] Bill Russell: [00:06:10] and I use the GPS to go everywhere and it's just, for me, it's taking away part of the cognitive load. Right. So it's, I, I could easily get on autopilot myself as I'm driving around and forget to make a turn. But if the autopilot's on, I'm not going to forget to make that turn. Cause it's gonna, it's gonna prompt me to make that turn.
[00:06:29] Anthony Chang: [00:06:29] Now [00:06:30] that's a good, insight too. So, but you still have to have human in the loop because if I just went with my GPS and automated driving out, for instance, as you recall, Newport Beach, there's a bunch of islands around here and it would have directed me, right off the cliff into the ocean, because if forgot that the car had to take, get on a ferry to go to, some of the islands Balboa Island.
[00:07:00] [00:07:00] So you have to have human in the loop. And that was a good reminder for all of us when I was seeing my GPS, heading the car right into the ocean, because it really didn't realize that it needed to connect me with the ferry. And I think that's the same in clinical decision. Right? So, or, or medical image interpretation. You don't want to have the human entirely out of the loop. and humans really need to be in the loop.
[00:07:25] Bill Russell: [00:07:25] Yeah, totally. All right. So, so let's the pandemic is top of mind. Let's talk about that a [00:07:30] little bit. Has the pandemic impacted positively or negatively the progress of AI?
[00:07:37] Anthony Chang: [00:07:37] I think both. I think if I were to give AI a grade for during this pandemic, it would be like a. B minus, C plus, and it's not the really the AI's fault. It's basically a manifestation of how badly we need much better data and IT [00:08:00] infrastructure and healthcare. I think we don't have those things in place, as well as a public health, strategy in place. Then the best AI is that, that this country has, is not going to help enough.
[00:08:14] So on the diagnostic side, I think we've really fallen short because the testing wasn't readily available still isn't. But using AI, for instance, you can apply machine learning [00:08:30] to the pool testing concept to increase the odds of finding a positive. person with the virus, although in some States at the positivity rate is so high, it doesn't even matter anymore. But I think you can use AI even on the diagnostic side. So on the therapy side, it's been really exciting because, now we can predict protein structure with just the genomic sequence using deep learning. We can also do a vaccine design and strategy with deep learning. [00:09:00] So all of these things are available now to help us with the therapy side, but it would have been nice to have the data in it, infrastructure, good enough so that the AI can really, take place. As a concrete example, it took us two to three months to realize that proning patients and we need a higher oxygen levels and their blood was just as good, if not better than mechanical ventilation.
[00:09:26] That should have been discovered within days and [00:09:30] weeks of this pandemic if we had real time analytics built in to data and it infrastructure in hospitals around the world. We should have just known that within days and weeks, but it took months. And it took months for intensive as the tech age out to learn that strategy.
[00:09:47] Bill Russell: [00:09:47] So the AI consumes data and specifically clean data, data that is ready to be consumed by AI. But I, I would think that the example you [00:10:00] just gave. I mean, we're getting telemetry data. We're getting all that stuff is relatively clean data. Why weren't we able to do stuff with that?
[00:10:09] Anthony Chang: [00:10:09] Well, maybe, maybe a clean data, but the data wasn't shared. So if you don't share data amongst the hospitals, then you're not necessarily going to have information and knowledge from that data. So we need, very badly we need a international consortium of hospitals to pool their data so that [00:10:30] when the next pandemic hits, then we're going to be much faster at coming up with insights about therapy. And this goes to not just, this goes to, medical therapy with random REM dense, severe, and all the other agents, antibody agents, so that you really need to pull that patient experience. So that the next patient will truly benefit from the cumulative knowledge that you built from prior patients.
[00:10:57] Bill Russell: [00:10:57] Yeah. And we're from a public health [00:11:00] standpoint, those, that data and that data architecture is almost not even at the starting gate for AI.
[00:11:07] Anthony Chang: [00:11:07] Yeah. Particularly in the US we should be embarrassed about our public health infrastructure is as global health backgrounds. And it's just astounding to me how, A few, second, third world country probably arguably have as good if not better public health infrastructure as we have. Now we have great expertise in global [00:11:30] health member autology and epidemiology. But when you don't have the infrastructure, just like you can have a great AI expert, but without a data infrastructure, you still, are disadvantaged.
[00:11:42] So I think one of the things I've said is, having good public health strategy coupled with good AI strategy is like having, protective gear as well as weapons when you fight a war and this case is the virus. [00:12:00] So you have to kind of have to have both, right. You have to have both good public health and data in it infrastructure with great AI technology, and that will be a very centered, just, Dion's
[00:12:15] Bill Russell: [00:12:15] Yeah. And the area where AI is going to get an A really is around the vaccine. I was, I was talking to somebody early on in this process and they said, don't expect a vaccine till well into 2021 [00:12:30] they said that that's just the normal flow, the normal site. It's going to take a long time before we have a vaccine. And, and this is somebody who worked in that business for the better part of 30 years. And they just said, that it's, it just it'll take them all of this year just to develop it.
[00:12:46] And it'll take them most of next year to test it out. But it looks like we now have, two candidates, Morderna and Pfizer that we could start seeing as [00:13:00] some general use and in December. And a lot of that. Or a bunch of that is due to the use of AI. Do you think?
[00:13:08] Anthony Chang: [00:13:08] Why? I don't know. I'm curious to find out after they release the vaccine, they probably don't want to say too much prior to the release, but I'm betting that they were using analytics to some degree to really accelerate the design of the vaccine. And, I think that's, it's hard to imagine that humans alone were [00:13:30] involved in that whole process. So many things you have to rule out, you have to get rid of non-effective possibilities pretty quickly. And the whole excitement about drug repurposing. I think we talked about this last time, is just using machine learning to find new ways of using drugs.
[00:13:47] I've already been approved for a different purpose and because the human mind can't, draw the parallels as quickly as computers can.
[00:13:58] Bill Russell: [00:13:58] the, the interesting thing about these two [00:14:00] drugs is they're both, genetic, drugs, vaccines, generic vaccines, as opposed to the more traditional vaccines.
[00:14:08] All right, we'll get back to our show in just a minute. I want to give you an update on the clip notes referral program. For those of you who don't know what clip notes is, what we have done is we create a summary of what goes on on each show and we get that into your inbox 24 hours after the show airs. So what you get is a. Brief summary of the show.
[00:14:27] You get bullet points as to what the high points were [00:14:30] with timestamps, and then you get one to four video clips. So if you just want to see that the highlights, the key moments of the show, you can just click, click on those, watch those and stay current. we did that in response to some of the things you guys were telling us, made it easier. Easier for you to, share the content of the show with your peers, with your staff, with, with others. And it's a great way for you to stay current, especially when you're busy and you can't listen to every episode. So we wanted to make sure that you have an easy way to stay up to date as to what's [00:15:00] going on in the show.
[00:15:01] As a result because you told us this is, has been such a great, service. We created a referral program because we want to get this in the hands of as many people as possible. And we can only do that through you. And, so what we did is we said, all right, let's, let's do a competition. Everyone likes competition.
[00:15:18] So we, we came up with some prizes. And some ways for you, to, refer this out to your friends, to your staff, to people within your organization. the prices are pretty straightforward. [00:15:30] Everybody who does a single referral will be entered into a drawing to win a work from home kit from This Week in Health IT
[00:15:36] If you get up to 10 referrals, you get a black mole skin notebook delivered to you that has this week in health IT logo embossed on it. I love the mole skin notebooks. I use them all the time. And then finally, the person who gets the most referrals will be given the opportunity to come on the show, to do a Tuesday news day episode with me. That's the episodes that I usually do with Drex DeFord. and you're going to get invited to come on and talk about the [00:16:00] news with me. It's really straight forward. A lot of fun. it's really conversational. We'll have a great time doing it. And, I hope that you are excited about this as excited as I am, to have people, put your name in for a referral right now.
[00:16:13] If you go out onto our website, hit subscribe. there's a new box in there. It says referred by if they put in your email address. you will be a credit for each referral that they put in there. we already have people that are, approaching the mole skin notebook levels, so gotta keep [00:16:30] pushing through and we have, we've had just a ton of referrals to this point. So we really appreciate all that you guys are doing. And, we hope that the content is helping you to stay current and helping you to do your jobs more effectively. So, with that in mind, let's get back to the show.
[00:16:45] I want you to project a little bit, so, let's project out. Let's make it a long time, 20 years from now, before the next pandemic hits. How could we experience the [00:17:00] pandemic better? And what things would you have in place in order for us to utilize data and utilize the technology in order to, creat a better way of addressing and handling the, the pandemic, say 20 years from now.
[00:17:18] Anthony Chang: [00:17:18] Well, I wish I could say that the next pandemic will be 20 years from now
[00:17:23] Bill Russell: [00:17:23] There they're coming. They're coming too quick these days. So I'm hoping
[00:17:27] Anthony Chang: [00:17:27] It could easily be next year. So [00:17:30] we need to, we don't have a 20 year timeline to come up with a better, data and it infrastructure. So in one of the dividends perhaps in this pandemic Bill is that I think we realized how inadequate we were was, the best, some of the best medicine that we have and yet we were paralyzed with lack of data and information from the healthcare system. So I think that's going to accelerate that development to build [00:18:00] better data, sharing data analytics across hospitals. And I'm already seeing that with some hospitals sharing medical image data to be interpreted.
[00:18:10] Yeah, because it's basically human versus virus right now. And, nothing bands people together more than an external threat, which this is, as a crisis. So I think humans are banding together. The traditional silos of academic centers are not nearly as, as high. [00:18:30] So I say collaborative efforts, as editor in chief of a journal called intelligence-based medicine, which is essentially AI clinical projects.
[00:18:39] I'm seeing collaborative efforts far more than even, earlier this year. So I think that's an encouraging sign. So we talk about collaboration between clinicians and data scientists, but I'm also hoping that collaboration will be among hospitals and health systems as well to really. put AI permanently on the, on the map for [00:19:00] improving health care.
[00:19:01] Bill Russell: [00:19:01] So, so let's talk about some of those projects. I mean, Med AI was doing some shark tanks, prior to COVID. What kind of projects are you seeing? What kind of interesting things are you seeing? What kind of startups are being. Formed out of the, out of the AI work that you, that you're a part of.
[00:19:17] Anthony Chang: [00:19:17] Yeah. And that was, sponsoring pitch events. And what I'm seeing more and more is, what the medical image interpretation was very popular as a [00:19:30] topic, just looking at CTS and chest x-rays to diagnose COVID. I don't think that particularly will make huge impact, even though that's been. Are they assigned to the stream to do that project?
[00:19:40] I think the much more practical aspects of delivering, COVID and post COVID care is where the action has been to some extent, and I'm encouraged by that. So by that, I mean, using machine learning to prioritize patients that need to be seen on that, that. we [00:20:00] are potentially in a position to get ready for reopening of a lot of clinics and centers. How do you prioritize those patients? Well, I think, using machine learning to, to put patients on a priority list will be one very good way to use machine learning because humans are just not very good at remembering all the details about all of their patients. And I have up to 10,000 patients I'm following.
[00:20:25] There's just no way I can remember. Everyone's sort of, priority, [00:20:30] sort of standing in terms of getting seen. So I think that's going to help workflow issues can be helped by as some sort of automation. So I think we're going to discover ways of using machine learning in a non. Sort of traditional way that we've been talking about in the last couple of years. So mainly getting machine learning into workflow rather than sort of more academic image projects.
[00:20:56] Bill Russell: [00:20:56] Yeah. So, the, it's interesting. I had a conversation [00:21:00] with LeanTaas earlier this year and they were talking about using AI, around the flow of patients into the OR and those kinds of things, and they were increasing the efficiency of the OR obviously that's where money is but now you go into a pandemic and they're actually applying that same logic across some other, some other areas to increase the overall throughput and efficiency. Olive AI has made some significant [00:21:30] moves in the RPA space. Those are all forms of, of, different types of AI. actually, can we go back as you, you did a, a great short segment in our last, interview, but it was almost two years ago.
[00:21:44] What are the different types AI that we're seeing being used in healthcare?
[00:21:52] Anthony Chang: [00:21:52] Well, I think, you can roughly describe AI and healthcare into three kind of categories. [00:22:00] One is assisted. So that's like your iron by wrap, vacuum cleaner. So they are robots and healthcare that can take a blood sample and do all the testing without humans involved and then automate the process and then send the results into the electronic breakfast. So that's assistant is a repetitive task that doesn't require a human to intervene or give you no oversight over. That's just a very kind of a robotic process. The middle category is [00:22:30] augmented AI in healthcare. So that's where you hear about Watson health getting involved with Memorial Sloan Kettering, even though didn't deliver as big as other than as everyone had hoped.
[00:22:42] But a lot of the analytics work that's being done with medical images and decision support are in that augmented category of what humans have to be providing some insight and experience and labeling for the, computer, data science to benefit [00:23:00] from that knowledge to make predictions basically make better predictions.
[00:23:05] And then you have on the other end of the spectrum, what's called, automated AI in healthcare. And that's, autonomous, it's also called, so that's where a algorithm can really make its own decision without the human in the loop. Now it sounds intimidating, but there are times that you really, don't [00:23:30] necessarily, have access to healthcare. So for instance, the IDX dr. software that is approved for autonomous AI in healthcare. Can interpret a fundus photograph for diabetic retinopathy. And, because that's based on a lot of ophthalmologists experience. So, but I still think with. Autonomous AI, the medical world is not quite ready to let these go entirely without [00:24:00] oversight.
[00:24:00] And I think it's good for a while to have oversight no different than any modern technology. And you just want to, have enough human experience, what the technology do feel more and more comfortable, but the human, acceptance overall for, AI is changing. I remember. Five years ago, even when I used to do an audience poll about, who would know riding the autonomous driven vehicle virtually no hands would go up and now more than half the hands will go up. [00:24:30] So I think we're sort of, increasing our. Are, accommodation of these technologies.
[00:24:37] Bill Russell: [00:24:37] Yeah. It's it's interesting. I want to go in two directions with you. One is there's a big debate around, ethics and bias, right? Because AI essentially is code behind it. That's being written by people and people have bias. People have yeah. I mean some prejudice and whatnot, [00:25:00] which can enter into the algorithms that are, that are utilized. where's that conversation going and, are we, I assume we're having that conversation. It's, it's, it's starting to, permeate, some organizations I've actually heard of some organizations setting up, groups that actually look at how they're bringing AI into their healthcare system and reviewing the algorithms for things like bias and, and those kinds of things.
[00:25:30] [00:25:30] Anthony Chang: [00:25:30] Couple of thoughts. One is, a lot of times the bias is not intentional. It's just that it was
[00:25:36] Bill Russell: [00:25:36] Oh, no. I've yeah, absolutely not. I, I didn't mean to imply that I'm just saying it's, it's just part of who we are, I guess.
[00:25:43] Anthony Chang: [00:25:43] Yeah. I think a lot of it is under intentional, but perhaps bias that we weren't realizing and sometimes it's. It's just appreciating, certain assumptions were not entirely correct when you look at the data. So I think one of the best things about AI in healthcare is that [00:26:00] as clinicians, we learn a lot about ourselves by and how we think by applying AI. So it's a reflection of how we have practiced in the past. Just as much as, discovering new things with machine learning. So point number one is, we need to be partners with being, in terms of being accountable for bias is not like the algorithm is biased. So that's the algorithm's fault where we should be 50- 50 partnership with algorithms in terms of, being [00:26:30] accountable for bias.
[00:26:31] And I think, the other thing is, as you have alluded to, if we have bias, even if it's unintentional, It's going to be just automated and perpetuate it. So it needs to be very careful that we don't do that as much as we can. So we kind of have to look at the entire workflow of these projects from how we label data.
[00:26:51] Right. Cause you can have bias just labeling data. From the labeling of data, curation and data through the algorithm process and then [00:27:00] outcomes, the prediction. And we still have to look at the prediction because maybe that's bias because somewhere along the line during the workload, things became biased. And, so I think it's, it's a tough problem, but I don't think it's a deal breaker when it comes to applying AI in healthcare. We just have to be. Very cognizant of potential biases and use our real world experiences to, to, to make sure that we have as little bias as possible. It may be impossible to get rid of bias [00:27:30] entirely.
[00:27:30] Because humans are involved in trying to figure this out and we may need the, the, AI to, to figure it out the bias in AI, ironically.
[00:27:40] Bill Russell: [00:27:40] Yeah. And I, the, the thing that's encouraging me is I am seeing health systems stand up governance around bringing AI into the organization, which means that it is it's first of all, it involves clinicians. There's physicians and nurses, who are a part of that process. A lot of times it was hard to get [00:28:00] the clinicians and nurses to be part of the, of the technology projects and they're heavily engaged. They see the value. and, and I just think the fact that they're standing those up, is a, it's a good sign, but, and, we'll see where that goes.
[00:28:16] One of the things I did want to touch on with you is AIMed is global. You were doing conferences in Asia, you were doing them in Australia. were you in Europe or
[00:28:28] Anthony Chang: [00:28:28] We are usually in Europe at least once in a [00:28:30] year.
[00:28:30] Bill Russell: [00:28:30] So you were all over the place. I'm curious, as you sort of look at this is, where, are the places where you're seeing the most innovation in AI? Is it the U S? Is it Asia? Is it other parts of the world?
[00:28:49] Anthony Chang: [00:28:49] Well I think based on emails and interests from different regions in the world, I can say, probably UK and parts of Europe are still very [00:29:00] engaged. China, for sure, Singapore and other Asian countries, Japan and Korea are also, and US obviously, but only, not the entire country, obviously, but certain pockets and then Canada Israel. So it's about a dozen places they're are particularly interested in engaged. we just were launched the American board of artificial intelligence and medicine, or ABAIM for, certification and knowledge, [00:29:30] starting next year. And I'm already getting calls from the UK, Australia, and China about bringing that process to their country.
[00:29:39] So that's hugely exciting for us. I didn't expect us to be international, so it really. But, we offer a two day review course and throughout the year, every month, and then we have a certification process that you can kind of assess your level of knowledge in this space. and with everyone, aiming to improve their knowledge. Of course. So the course is [00:30:00] basically based on the textbook, the textbook is a little bit daunting if, even though I think it's relatively straightforward reading for any clinician. But, so we kind of divided the book into shift the educational modules that we review for the review course.
[00:30:19] Bill Russell: [00:30:19] Oh, what's, what's the pandemic been like for you? This is my curiosity thing here. Cause I don't think I knew of anyone who was more busy than you. prior, prior to the [00:30:30] pandemic, and the pandemic slowed us all down. Right? I mean, there was there w we weren't traveling as much. I think I've only been on the plane, a plane twice this year, and it would normally be about 40 or 50 times by this point of the year and you would have amassed a ton of miles. What did you do with that time? Because you're not one to sit around much, what is it, what did it look like for you?
[00:30:55] Anthony Chang: [00:30:55] Well, of course, it's a great time for self-reflection, right? Because we have, [00:31:00] are forced to have a lot of time alone or with your family. First of all, the most important dividend for me during this pandemic on the personal side is. As a data scientist I calculated out the increase in percentage wise, the number of interactive hours I have with my two girls. As you recall, there were a little and, oh, they're bigger now they're five and seven now Bill.
[00:31:24] So it's really a great time to spend time with them. [00:31:30] It's almost four fold increase in the interactive time number of hours. And that's been just amazing for me that. in the middle of this pandemic, we need to look for positive dividends. And one positive dividend is we've gotten closer than ever before. And just the sheer volume time that we're together, 24 seven literally, the bad experience has been great. and then because of [00:32:00] traveling less and having, less traveling just locally, but right, because I spent an hour and a half probably going back and forth to the hospital, gained the time gain we've, launched the American board, which no one thought was possible in six months, but we did, we have, the Bookout launch and then I'm following that up with an AI and cardiology book cause I'm a cardiologist. So that's launched. we have, a video [00:32:30] series and production. So it's not like, because I have a lot of time now I can do all that, but it just, you direct time to projects that would have been harder when you're, more active traveling.
[00:32:43] Bill Russell: [00:32:43] So at CHOC you talked about not commuting in as much. Are you guys using, in hospital telehealth solutions to do rounds and those kinds of things.
[00:32:53] Anthony Chang: [00:32:53] Yes. Where I'm seeing my patients virtually now I miss them terribly. As a pediatric [00:33:00] cardiologist you are used to hugging your patients and your family members, but now there's not that. So I see patients virtually and to my surprise, I think a lot of patients or families are actually quite happy with that arrangement doing, especially during pandemic, because they don't like the risk of going in either on not just myself, but, so, but it's not the same, but I think in the meantime, it also taught me that perhaps a third, my visits, but easily be done [00:33:30] to a telehealth and not have to inconvenience, the family needs to come in, even without the pandemics. Because a lot of my patients are handicapped and they get Durham wheelchairs, and it is a big deal for them to get in the van and come over and see me. So I'll go to the clinic and see me. So I've learned to, it's a different practice of medicine, as you can imagine, you have to trust your eyes even more than before. so I think, that's worked out, I think just fine for almost all the patients at dams.
[00:34:00] [00:34:00] Bill Russell: [00:34:00] Yeah. I mean telehealth will be one of the things we look at and we'll have fundamentally changed. I remember sitting across from some physicians who were looking at me and say, I'm never going to do telehealth. You can't do all this other stuff. And those same physicians are, I'm sure. Using telehealth just like you probably a third of their patient visits.
[00:34:20] Anthony Chang: [00:34:20] We're also learning a very exciting area, which is using AI in wearable technology. So, all the wearable devices are [00:34:30] not very helpful unless you can either embed or use AI for the data that's going to generate. So, because humans are just not gonna be able to keep up with that streaming of data. So in a way it depends on it also gave us an opportunity to explore that area.
[00:34:46] Bill Russell: [00:34:46] So what's next for AIMed what's what does AI med look like in 2021?
[00:34:50] Anthony Chang: [00:34:50] Well, in 2021, we'll stay virtual. We're going to have a big blowout events, hopefully in 2022 in January. So we're looking forward to that. [00:35:00] Rekindling friendships and, a screed of core. we're going to stay virtual. We're going to have a lot of activities. We're going to go into a lot of subspecialties. As you recall. Last two, two years ago, we just started going into clinical set of specialties, like cardiology and radiology, but we're going to expand into 10 or 12 more.
[00:35:21] We're also going to, have, a more active, ecosystem and [00:35:30] subsystems. So we launched something called AMM connect as a way to help clinicians and companies and hospital systems and academic centers to connect in a certain category of AI. So that's been very helpful and successful. So. just getting all the stakeholders together to talk about the problems in AI and healthcare has been tremendous. So, they're going to be, pretty active and, and busy next year.
[00:35:58] Bill Russell: [00:35:58] Yeah. So you're, you're [00:36:00] expanding the community even as you're not getting together, the community is growing and the connection is growing. If people want to be involved in the stuff you're doing, you guys do some podcasts as well, right?
[00:36:10] Anthony Chang: [00:36:10] We do quite a few webinars. Yes.
[00:36:13] Bill Russell: [00:36:13] Do you do webinars? Okay. So if people want, if people want more information, they want to get connected. What's the best way for them to do that,
[00:36:20] Anthony Chang: [00:36:20] They can just go on our website, which is, AI dash med dot mail and has all the events and activities. [00:36:30] there's also now the American board, ABAIM.org. That's for the American board and they can sign up for the courses as well as, the certification process that they want to get their, group certified, starting next year. That's a good way to get certified and have a crash course in AI. If you don't want to go back to the school like I did for four years, a two day course really kind of gets you up to speed very quickly.
[00:36:59] And [00:37:00] then we have a, medical intelligence society or MIS that's mainly for clinicians, but it's open to everybody. Just like everything I do. And that's where the clinicians get together and talk about their clinical projects. from the academic perspective, when we talk about journals and journal articles. So it's particularly good for clinicians, but it's open to everybody. So. We have a lot going on trying to keep all these things going during pandemic, [00:37:30] but I'm grateful to all the teams there are behind all of these, major initiatives.
[00:37:35] Bill Russell: [00:37:35] Yeah, no, it's fantastic. You've put together quite the community if anyone's listening to this and they want to hear the really cool story of how you got into AI and going back to school, what they should do is go back and listen to the archive cause I, that really was, I enjoyed that episode. I enjoyed sitting up there. We were at the innovation Institute down in Newport, overlooking the ocean and, yeah, I can't wait til we get back together again and [00:38:00] I'll come back to Southern California. We'll find a spot looking at water and we'll have the next conversation. Maybe 20, 22, but, I'm looking forward to it.
[00:38:10] Anthony Chang: [00:38:10] Yeah. Hopefully I'm over this pandemic and not. in the middle of another pandemic. So I think there's a sense of urgency for us to really get our act together that humans are particularly caught off guard with the magnitude of the buyers and their capabilities. And, you can learn from viruses too, even though [00:38:30] they're technically not even a living bill, right. They're just, Some, the DNA sequence. They're not technically living, but you know what? They're very well-behaved they work as a team. They don't have a centralized leadership that can get in the way. They just all function with one goal in mind, in fact kill humans and duplicate. So,
[00:38:53] Bill Russell: [00:38:53] And then you're saying we could do that same thing.
[00:38:56] Anthony Chang: [00:38:56] I think we can learn from there. What's good [00:39:00] about the virus, which is, technically a very well functioning, complex adaptive system. They adapt to changes in environment. They work in unison as a. As a team. And I think humans can learn from that. And hopefully we will, we will be able to learn from this virus.
[00:39:17] Bill Russell: [00:39:17] Totally. Anthony, thank you again for your time. I really appreciate it. And it's always good to see that Southern California sun behind you.
[00:39:26] Anthony Chang: [00:39:26] Yes, and we, very, very much looking forward to Bill [00:39:30] having another chat in person. So thank you very much for the opportunity.
[00:39:35] Bill Russell: [00:39:35] Absolutely that's all for this week. Special. Thanks to our channel sponsors, VMware, StarBridge Advisors, Galen Healthcare, Health lyrics, Sirius Healthcare, Pro Talent Advisors, HealthNEXT and McAfee for choosing to invest in developing the next generation of health leaders. We really appreciate their support. Don't forget to sign up for clip notes. send an email, hit the website. we want to make you and your system more productive that shows the production of this week in health IT. For more [00:40:00] great content, check out the website this weekhealth.com. Check out our YouTube channel. We continue to modify that and make that better and easier to find things that you are looking for there. please check back every Tuesday we do news day, every Wednesday, we try to do a solution showcase every Friday. We do, Interviews with industry influencers and we will continue to do that through the end of the year.
[00:40:20] And then we have some interesting things lined up for the new year. And I can't wait to share those with you and we will start sharing those with you here shortly. So, [00:40:30] thanks for listening. That's all for now.