The AI Revolution in Healthcare: AWS Unveils Cutting-Edge Tools

0:00The wait is over. Dive into Audible's most anticipated collection, The Best of 2025. featuring top audiobooks, podcasts, and originals across all genres. Our editors have carefully curated this year's must-listens from brilliant hidden gems to the buzziest new releases. Every title in this collection has earned its spot. This is your go-to for the absolute best in 2025 audio entertainment. Whether you love thrillers, romance, or nonfiction, your next favorite listen awaits. Discover why there's more to imagine when you listen at audible.com slash best of the year. I am always really excited about the use of AI in healthcare.

0:44And today on the podcast, we're going to be talking about two pieces of AI news. One, that AI is helping in detect breast cancer and helping reading x-ray screens there. And the other is a bit of news out of Amazon and AWS on how they're helping with healthcare. So let's jump into it. I think the first one I want to cover today is the fact that there was a study done showing how AI could help safely read mammograms. So the study that they did on this involved around 80 ,000 women from Sweden with an average age of 54. And the results showed that AI could, I think, pretty substantially reduce the workload of radiologists without increasing the rate of false positives, right?

1:24This is the big thing that they're worried about. if the AI can perhaps detect more, but it also has more false positives, it may not actually be saving time. So the idea here is that the AI would actually be able to reduce the workload of radiologists. So the study had half of the scans assessed by two radiologists, which I think that's actually the standard. Then they had the AI help with looking at the scans. So after they did this, I think the study had shown that AI supported screening, right? So they actually used AI supported tools in the screening. I think they did actually have a radiologist, one radiologist review afterwards.

2:00But the AI supported screening identified 244 women with cancer. And then they had, you know, compared that to 203 women that were just the standard two radiologist screening. So the AI was actually able to identify over 41 extra people that had cancer and the false positive rate remained consistent at 1.5 % across both groups. So these were actual people that had cancer that probably, you know, that could have been missed without this AI technology. So I think most notably, researchers believe the implementation of AI could potentially half the screening workload. So a 44 % reduction in the screen reading workload of radiologists was observed with, you know, 36 ,000 fewer screen readings by radiologists in the AI-supported group compared to those who had received standard care.

2:55Dr. Christina Lang is the study's lead author from Lund University in Sweden, and she commented on the results. She said, the greatest potential of AI right now is that it could allow radiologists to be less burdened by the excessive amount of reading. So I think despite the, you know, know promising results she also cautioned that the findings alone are not enough to confirm that ai is ready for full implementation of you know mammogram screenings and she said quote we still need to understand the implications on patients outcomes especially whether combining radiologists expertise with ai can help detect interval cancers that are often missed by traditional screening as well as the cost effectiveness of the technology the nhs already is looking into the implementation of this kind of technology into its breast screening programs and it has voiced optimism about the study.

3:42I believe an NHS spokesperson said the NHS is already exploring how AI could help in breast screenings by enabling complicated image analysis very quickly and at scale which if proven effective could in future help speed up diagnosis for many women detect cancers at an earlier stage and ultimately save more lives. So I think though more studies are needed to kind fully comprehend the potential of AI and breast cancer screenings. I think these preliminary findings here are a really impressive result. And I think that AI is going to play a really critical role in kind of easing the workload of radiologists and in many, many different types of doctors in the near future.

4:24So I think that is really, really exciting. The second story I wanted to cover today is a way that Amazon and AWS are helping with healthcare. So AWS has announced two new services that I believe could transform kind of clinical documentation and image management. So announced at their annual summit in New York, AWS is set to add these to its suite of cloud-based health services. And one of the services is called AWS HealthScribe. So this is a generative AI tool, which is equipped with text-to-speech capabilities that creates conversation transcripts. So it uses natural language processing to essentially segment the content into categories such as key takeaways, um reason for the visit and history of the present illness and it does this to you know essentially create these summer summaries so clinicians can essentially review and finalize these transcripts and summaries into their electronic health records i believe this is a really no-brainer um kind of implementation of ai and i think we've seen a couple other um ai players working on similar tools but this is a hipaa eligible tool and offers a single api which is giving health system control over kind of data storage and maintaining encryption during visits and at rest.

5:39And I think that's a really big part of this. Obviously, this has to be HIPAA compliant. According to AWS, clinicians can choose, you know, whether they want to store transcripts and preliminary clinical notes and kind of retaining ownership of the content, which is really important. So the second service that AWS is launching is called Health Imaging. And this is a new medical imaging service that handles data at petabyte scale. So it can potentially save organizations money when they transition from self-managed infrastructure to DICOM ingestion. So the service imports individual DICOM P10 files as image frames and organizes them in image sets with consistent metadata at the patient study and series level.

6:20So companies such as Wake Forest, Baptist Health, and Philips are already leveraging the service. And Wake Forest, Baptist Health is urging it for research and education across its enterprise, while Philips plans to optimize workflows and accelerate the time to diagnosis and treatment using AWS health imaging. So Bratton Sa, who's the vice president of machine learning and AI services at AWS, really highlighted the kind of broader goal of these services, saying, quote, our healthcare customers and partners tell us they want to spend more time creating innovative clinical care and research solutions for their patients while spending less time building, maintaining, and operating foundational health data capabilities.

7:04So the AWS developments really align with an emerging trend of AI's increasing role in healthcare. I think last month, AWS launched the AWS Generative AI Innovation Center to kind of further enterprise innovation. And they had more than 100 ,000 customers already using AWS AI and machine learning services. So I I think these new additions are really poised to make a fairly significant impact in the field of health care and really improve patient care and reduce the burdens on, you know, health care professionals overall. Right. This doesn't necessarily replace doctors with this AI or radiologists or anything else, but it really serves to help and help them do their jobs better and, you know, potentially replace some tasks they have to do.

7:50radiologists, you know, looking at x-rays. It's interesting because in all of the studies and all the research, I think people are really, they're really sort of cautious and they're like, oh, don't worry, like this isn't to replace a radiologist. This is just for radiologists to, you know, help reduce their workload. And it's kind of funny because like, if it really was as accurate, if it really was more accurate than a radiologist on a consistent basis, then why wouldn't you just replace a radiologist with this right but they don't want to say that because they don't want radiologists to freak out or people to freak out and say like the AI is taking jobs no one wants to say AI is taking jobs they just want to say the AI can help you do your job better because that's I don't know that's that's a pretty good narrative right less people are afraid but I think at the end of the day with this and with a lot of other technology if it can do the job better than I'm not saying this is good or bad either by the way I'm just saying this is the inevitable outcome is that if it can do the job better than a regular radiologist, it will be what is used.

8:48It will probably replace radiologists. And if it does a better job, I'm not saying that's necessarily a bad thing either. So I definitely think that's something that's interesting. And it's going to be very interesting to follow how these tools improve. And when they improve significant beyond radiologists, what role they'll play, what role radiologists will play in this and what that will look like in the field in the future.