Reddy.AudioPodcast.Abdou.022125.NURS844

Dr. Varudhini Reddy explores the intersection of artificial intelligence (AI) and healthcare with industry expert Meena Abdou. AI is transforming healthcare by improving social well-being, analyzing data, and making personalized treatment plans. Software engineering plays a crucial role in integrating AI algorithms into clinical workflows. Challenges include data management and privacy concerns. AI has the potential to predict diseases and personalize medical plans, but it requires data and societal acceptance. Hi, my name is Dr. Varudhini Reddy and I would like to welcome you to today's episode on artificial intelligence and healthcare. During this audio podcast, I will explore the intersections of artificial intelligence and healthcare with industry expert and vice president at JPMorgan, Meena Abdou. Artificial intelligence is changing the world at a fast pace and transforming the delivery of healthcare. Today, I will delve into how artificial intelligence is transforming healthcare investment opportunities, data analytics, and future trends. Artificial intelligence is a technology that allows machines to learn, reason, and act like humans and has the potential to improve social well-being. AI systems learn through analyzing large amounts of data and use algorithms to identify patterns and relationships in data. Predictions are made based on these patterns and currently, AI is revolutionizing various industries. Healthcare is no exception. From predictive analytics to personalized medicine, AI is enhancing healthcare delivery, proving operational efficiency, reducing physician burnout, and driving creativity and innovation in the healthcare sector. Today, I will highlight key aspects of AI in healthcare through the lens of finance and software engineering. Software engineering plays a crucial role by developing applications that enable the integration of AI algorithms into clinical workflows. This allows for rapid analysis of patient data, improved diagnosis and personalized treatment plans, and more streamlined healthcare operations. The potential of AI may lead to better patient outcomes and more efficient healthcare delivery. In addition, finance plays a crucial role in this transformation by improving healthcare financial operations and supporting the commercialization of AI-driven healthcare solutions. On that note, I would like you to introduce yourself, Meena. Sure. My name is Meena Abdou. I graduated the University of Delaware with a master's degree in software engineering with a concentration in artificial intelligence and information systems processing, and I've been working as a software engineering lead at J.P. Morgan for 12 years. That's wonderful. One of the questions I wanted to ask you is, what do you feel digital health is, and what does it mean to you? Yeah, so there are two ways to interpret the phrase digital health. The first is the health of your digital information, which is not the context of this conversation. The second is healthcare for humans delivered digitally, and that is becoming really important because more and more of our lives are controlled or delivered digitally. Most people communicate through their electronic devices. We consume media over our electronic devices. We now even get our food delivered through our electronic devices. It only stands to reason to expect that pretty much every facet of our lives will be delivered digitally at some point or another, and that includes healthcare. And as part of the role of AI, it's really a continuation of what has been happening in the e-commerce space. Everyone knows that when you go to Amazon.com and you look for something, it shows you a ton of recommendations, and if you buy that something, it will continue to show you more recommendations for similar items to buy. And if you comment on something, or if you like something, or if you share something, it's tracking all of that data, and now it's a common phrase in the industry that Amazon and Google know more about you than your doctor. And the role of AI in healthcare is to kind of bring healthcare up to par with the e-commerce space, to where it can actually know your health statistics, know what conditions you have, know how those compare to people that are in a similar physiological makeup as you, and it can make recommendations specifically tailored to your healthcare needs. And so it's bringing the digital delivery of healthcare up to speed with the digital delivery of goods and services. Thank you so much for your response, Lina. That is a wonderful summary of some of the benefits of AI and how it transforms healthcare delivery today. What do you believe are some of the existing challenges with artificial intelligence in healthcare? So number one is definitely the data management aspect. In order to have AI, you need to train it on data. And in order to get data, you need people to agree to release that data. And anywhere there's going to be a lot of data, there's going to be a threat of hackers or scammers trying to steal that data. And more importantly, there's also the threat that that data may be shared without your permission. And so, as you can imagine, we already have problems with identity theft, credit card numbers, social security numbers, and this is a big cybersecurity nightmare for a lot of corporations, and governments get involved and create regulations. If we now have this same level of scale with healthcare data, it can get even more risky. And most people are not comfortable with their healthcare information being shared with lots of third-party companies. And also, you never know if that data gets hacked and published on the Internet what the effect will be. The other challenge is what you could call the creepy factor, where most people are comfortable going to a doctor's office and talking to another human being and hearing recommendations or advice and so on and so forth. But when it's a robot telling you that, or it's a computer telling you that, some people get really freaked out by that. It becomes scary. It reminds them maybe of certain dystopian movies that they might have seen in the past. And so, how is the best way to deliver the information to the consumer? And then finally, how do you make sure that, again, privacy, if you are doing analyses across multiple data sets, across multiple populations, how do you make sure that if you accidentally send somebody healthcare information that doesn't belong to them, that you successfully recover from that? So, those are the biggest challenges right now with AI and healthcare. Great. Thank you so much for discussing some of the challenges we face today with artificial intelligence and healthcare among all aspects of the industry, and as well from the consumer perspective as well. Can you discuss any specific artificial intelligence algorithms or models that you've experienced or have shown promise in healthcare applications? So, I think this question could be reworded a bit. In the world of AI, we don't really talk about algorithms like in the world of broader math. A model is a specific instance of a technique that was used on a specific data set. And so, it's not necessarily that one model or another is useful and that there's no such thing as a generalized model. It's really more about the technique or the type of strategy you're using. If you are analyzing, let's say for example, imaging, healthcare imaging such as x-rays or CAT scans, you will need one type of technique to analyze those and to predict and have an AI system predict what the image is showing you. If you are analyzing healthcare study data, either through reading research papers directly or by taking all their data and converting it into numbers and then crunching the numbers, you need another strategy. If you're analyzing clinical data, like taking patient history, how patients communicate, how they refer to specific symptoms to try to diagnose what they actually have, then you need a different strategy. There are tons of different strategies available. The latest one that everyone knows is large language models. And these are very helpful for language-based interactions, whether it's a lot of text or a lot of speech. And they're actually not very useful for numerical processing. There are more efficient strategies or techniques for numerical processing. More than likely, the way that humans will interact with AI is just by talking about it or talking to it with natural language. And so LLMs will be most important in that area. And what role do you believe artificial intelligence plays in predictive analytics and personalized medicine in the healthcare sector? So I think it will play a big role depending on if we can get past the challenges that I spoke about earlier. And the reason why it's not that useful right now is that we still actually don't have a large group of, you know, a critical mass of people who are willing to share their own healthcare data in real time to a single company or maybe multiple companies that would allow them to build actual useful AI models that could allow them to predict things. But it is possible. The technology is there. It's just the will and the psychological effects of accepting to let the technology be used in that way is still not there. Obviously, what it can do, it can do a lot of things. It can predict if you're going to have, you know, genetic diseases if you, you know, meet with a certain person. It can predict your risk of developing certain chronic illnesses based off of your diet. It can predict based off of your environment, any environmental factors, psychological factors, other biological factors. And so it can personalize a medical plan that's tailored to every specific individual. But again, it needs the data to do that. And it will need the processing power to do that as well. And we're going to need to agree as a society that we want that before it can happen. Awesome. Thank you so much for joining us today, Meena, and sharing your insights on artificial intelligence and healthcare. I really appreciate your time and look forward to hearing more about exciting developments in this field. Thank you so much.