Dr. Jigar Patel

speaker

105 appearances 1 recordings 1 series first heard Dec 2023 last heard Dec 2023

Dr. Jigar Patel’s voice in public audio — every appearance, attributed to the second.

Trend

recordings per month · last 12 months

No recordings in the last 12 months.Older appearances are listed below; set an alert to hear about the next one.

Appearances

newest first · ▶ plays the moment

Health Chatter · AI in Healthcare · 15 Dec 2023 podcast

Yep. One of the first ways is artificial intelligence and generative AI in particular is good at doing summarizations. It's taking a lot of different sources and pushing together in a way that provides a breadth and depth of information that a static source might not have.

Now, the other thing that's interesting about generative AI is that you can also use it in chat GPT in particular is very good at this. In that it can actually tune up or tune down the literacy level of the output. So, giving me somebody that's graduate school educated, yourself as well, and saying, okay, normal patient education comes at a fifth to eighth grade reading level. Okay.

For me, that's like, I don't even bother looking at it, right? Because I know way more than that. But... If I do that, can I tune it up and give me more information at the level I'm going to understand is really quite powerful. And similarly, on the downside of that, and one of the things doctors do well or badly is explaining to patients, right?

And if you can tune it down to their level, because I often, one of my colleagues here calls me the chief explaining officer because I do a lot of explanations around this stuff. And I do it, but I break it down so that I don't try to be condescending, obviously, but I want to make sure people understand the basic building blocks of the concepts we're talking about.

And that's hard for physicians. But we're trained in the fancy words we've learned, right? But when you're talking to a patient, you start throwing fancy words, they don't get it, right? You lose them quickly. And the good clinicians then figure out a way to make it simpler for those that, and getting to the level they're at. AI can do that on the fly, right?

in a way that's unique and fast that we can't imagine doing and we can do more uniformly. Similarly, it also, people, there's been studies and other things, people actually think the responses out of AI are more empathetic than their providers. So it gives you literacy, it gives you empathy, it gives you different things that you're like, wow, didn't think about that at all.

It can, absolutely. So it's very powerful that way. And there's ways to have it do things that take work from a human mind perspective. It can do instantaneously.

Yeah, I mean, we're already, even without AI, we're already there, right? Social media, which is, there's a fair amount of AI in social media algorithms, machine learning sort of thing. We're already there. And so it behooves the public to understand the technology and how it can manipulate you without you realizing it, right? And there are companies that are using it to sell you things.

There are companies that are using it to grab your attention to things. It's the proverbial rabbit hole. AI can only make the rabbit hole worse. And if you're in a certain mindset, going and following the rabbit down the hole becomes easier and easier. We're already in a place where that's easy, right? It's going to get worse. So we have to educate people on the implications of this. I'm

And this goes to training of medical students, right, is a core example. I'm worried about giving AI to medical students because they don't learn to think like doctors, which is really, really important, first and foremost, as an example. We have a technology where we can record a doctor-patient conversation and then create a note for the doctor. I'm the first.

I told my product management team, do not put this in the hands of medical students. because you're synthesizing something that their brain needs to hardwire before they become physicians. So it becomes problematic that way too, in that we can, the hard work of becoming a doctor or the hard work of being an expert in anything can go away.

And that's, that's detrimental to us as a society and as a human individual. Sure. So let me, let me,

being an EMR electronic health record guy for most of my career, we had this problem with EMRs in understanding the technology. When I started, the joke was, there are some docs you're going to have to train to use a mouse. Exactly. That still exists. Right. Exactly. Not as much now, but 15 years ago? Absolutely. We had to worry about that. Yeah.

So it's going to be, I think it behooves every professional organization that's out there certifying their physicians on continuing medical education to have informatics and AI type conversations and training for those professionals. So it impacts them specifically from a trusted source that's understanding what they are.

It's more than just understanding the evolution of disease and the new tests and the new medications and those things. It's got to be this also. So on that front end, that has to be it. It has to be colleagues that are knowledgeable, CMIOs and others like myself, also being talking to folks like that's on health chatter to give them the viewpoint because I'm steeped in it every day.

So it's got to be multimodal in its approach. It's got to be, we got to blanket that across the board. Now, as we move down the spectrum from seasoned to younger individuals who have their MD, there's an advantage to helping them with things with AI, but they have to know AI is helping them. In a way that is different.

So the thing I tell people from a design perspective, which is really hard, is how do I let you know the validity of the thing I'm suggesting to you? Or how do I let you know this was created by AI? How do I give you indicators so you realize this is not another human, this is not something that was already there.

It was something that was created out of thin air, so to speak, and that's not entirely true. The knowledge and the exposure from a usability user experience perspective has to be there kind of in the workflow as well. Then as you think back into medical education, we still haven't cracked the nut on basic informatics education from a medical education perspective.

I, as a pathologist, it's part of pathology. Pathology was the first set of professionals along with radiologists that were using computers because of volume, because of those technologies. We had to be there. So I was taught that in my training. And I actually, in training residents in pathology, that was my job as well. I was the informatics guy. I was teaching them about informatics.

So we have to get back all the way into medical school and say, okay, here's your informatics course. Because Medicine is information, right? Treatment of patients is information. It has to be more than information. It has to go from data to information to knowledge in a way that's clear and open and clear cut to that person that's learning it. And they got to have the underpinning.

Showing 21–40 of 105 · page 2 of 6