Best of LinkedIn: Health Tech CW 42/ 43

00:00:00: This episode is provided by Thomas Allgaier and Freeness, based on the most relevant LinkedIn posts on health tech in CW.

00:00:06: forty two and forty three.

00:00:08: Freeness equips product and strategy teams with market and competitive intelligence to navigate the health tech landscape.

00:00:14: Welcome to the deep dive.

00:00:16: This week we're digging into the health tech activity that really stood out on LinkedIn over calendar weeks.

00:00:20: forty two and forty three.

00:00:22: And what really jumped out at me.

00:00:24: It feels like we've moved past the purely theoretical stuff.

00:00:28: The big theme seems to be practical integration.

00:00:30: You know, actually embedding intelligence, embedding automation right into the clinical workflow now.

00:00:35: That's spot on.

00:00:36: The focus is definitely shifting towards, well, operational results.

00:00:40: Real ROI.

00:00:41: So our mission today is to unpack the top, let's say, four clusters of innovation.

00:00:44: we saw driving that.

00:00:45: We're going to talk about AI helping out frontline staff, a huge issue with workforce shortages.

00:00:50: Yeah.

00:00:50: And also look at the pretty radical upgrades in imaging.

00:00:53: And maybe most importantly, how the industry is trying to build a better foundation of data trust and making systems actually talk to each other, interoperability.

00:01:01: Okay, let's dive in.

00:01:02: Theme one.

00:01:03: AI in care delivery.

00:01:05: It's obviously a massive topic, but like you said, the compelling part now is how AI has gone from being maybe a nice to

00:01:12: have to

00:01:13: really an essential tool, especially for easing the burden on staff.

00:01:17: And yeah, if you look at where the real pressure points are, it's often nursing.

00:01:21: There was a key post focusing on how Oracle Health EHR is working to empower nurses.

00:01:26: They're using voice first AI.

00:01:29: Voice first.

00:01:30: Yeah, to make documentation easier.

00:01:32: Things like AI-powered patient summaries, letting nurses use their voice for charting, like putting in discrete data, vital signs, that kind of thing.

00:01:38: That sounds crucial.

00:01:39: I mean, nurses spend so much time just on admins, staring at screens instead of with patients.

00:01:43: Cutting down that documentation time, that must directly help with satisfaction, retention, things every health system is worried about.

00:01:50: Exactly.

00:01:51: The aim, like Oracle Health's Seema Verma talked about, is less screen time, more patient care time.

00:01:57: Simple as that.

00:01:58: But it's not just admin.

00:01:59: AI is tackling.

00:02:00: We also saw it moving into some pretty high stakes clinical settings.

00:02:03: Like the ICU.

00:02:04: That needs incredibly fast, precise decisions.

00:02:08: Elk and Earl from T-Systems was discussing AI in the ICU.

00:02:12: things like ventilator weaning, super critical.

00:02:15: What exactly is AI helping with

00:02:17: there?

00:02:17: Well, this is where the prediction side really comes into play.

00:02:20: The AI looks at, you know, historical patient data, but also real time information like vital signs to predict the best time for extubation, taking someone off a ventilator.

00:02:29: Okay.

00:02:30: We saw sources mentioning prediction accuracy over eighty percent for finding that optimal timing.

00:02:35: Now obviously it doesn't replace the doctor or nurse, but it gives them a much stronger basis for their decision.

00:02:40: It could mean safer choices, potentially shorter ICU stays, better use of beds.

00:02:44: That eighty percent accuracy is a real tool against the risk of having to re-intubate someone.

00:02:49: And that hits the bottom line costs and outcomes.

00:02:52: But AI is also getting into that huge operational tangle with payers and the whole patient journey, isn't it?

00:02:57: Ganesh Kamath mentioned seeing companies automating that whole process.

00:03:00: Yes, absolutely.

00:03:01: The trend is definitely automating that whole administrative pathway.

00:03:05: Think about prior authorizations, always a headache.

00:03:08: We saw CoheerHealth automating this stuff at a massive scale, like over twelve million requests a year of their handling.

00:03:15: Yeah, that's a million.

00:03:16: Yeah.

00:03:16: And the reported outcomes are pretty transformative.

00:03:19: Things like up to seventy percent faster access to care for patients.

00:03:22: Yeah.

00:03:22: And maybe even more eye catching.

00:03:24: Forty percent lower admin costs for providers and payers.

00:03:27: Hold on.

00:03:28: Forty percent lower admin costs.

00:03:29: That's not just tweaking things.

00:03:31: That's potentially disrupting the whole model.

00:03:33: It really is.

00:03:34: Yeah.

00:03:34: And it goes beyond just billing.

00:03:36: There was another post talking about Emirates Health Services working with Oracle Health using AI specifically to cut down hospital readmissions.

00:03:45: So you see AI moving from just you know, documented care to actually helping ensure quality and monitoring patients after they leave.

00:03:52: The goal seems clear.

00:03:54: Put in smart automation to cut costs and risks at the same time.

00:03:58: Okay, so that speed and precision, like in the ICU example, well, it needs equally advanced tech underneath it, which brings us nicely to theme two.

00:04:06: Precision diagnostics, imaging, and novel computing.

00:04:10: This is where the basic hardware gets a serious upgrade.

00:04:12: Well, cut your eye in speeding up clinical analysis.

00:04:15: Yeah, this is where things get wild, I think.

00:04:16: The crossover between computing power and imaging.

00:04:19: Quantum computing, it's not just theory in a physics lab anymore.

00:04:22: Theodore Saz pointed out a breakthrough using quantum computing integrated into medical imaging.

00:04:26: It boosted processing speed by twenty-five percent for diagnostics.

00:04:30: Twenty-five percent faster.

00:04:32: And crucially, they said this was without losing accuracy.

00:04:36: diagnostic confidence stayed above ninety-five percent.

00:04:39: Just think about what that means.

00:04:40: Clinically faster results, more patients through the scanner in busy departments, quicker answers for urgent cases, that kind of computation just changes the speed limit for everything.

00:04:50: And that speed feeds right into new machines.

00:04:52: We saw new hardware using built-in AI.

00:04:55: A Koa Latu showcased the CT-五-three-hundred.

00:04:58: It uses AI to boost the clinician's confidence, but also gets eighty-five percent less noise and uses eighty percent less radiation dose for the patient.

00:05:05: Lower dose, higher confidence.

00:05:07: That's becoming the standard demand now, right?

00:05:08: Definitely.

00:05:09: And similarly, Fotis Lacha has talked about GE Healthcare's Cygnus Sprint Elite MRI.

00:05:14: It uses AI specifically tuned for tricky areas like cardiac and oncology MRIs.

00:05:19: Making those complex images, which often had artifacts, more reliable and faster to get.

00:05:23: I think

00:05:23: the really big strategic jump here, though, is towards genuine precision medicine.

00:05:28: Yolbakus discussed how AI is tackling this massive complexity problem in something called theranostics dosimetry.

00:05:35: Now, for listeners, maybe not deep in this, theranostics dosimetry is basically figuring out the exact personalized radiation dose when you're treating cancer using methods that combine diagnosis and therapy.

00:05:46: It's like a super customized digital plan for each patient.

00:05:49: It's an incredibly complex calculation, yeah.

00:05:52: And what AI is doing is using sort of simplified imaging combined with digital twins, these virtual models of the patient, to personalize those dose calculations.

00:06:01: It makes a level of customization practical that just wasn't before.

00:06:05: And this precision, it links straight to improving equity, doesn't it?

00:06:08: Carly Yoder highlighted that huge issue where, say, rural heart patients might have to drive for hours just for a basic scan.

00:06:14: Her point was really powerful.

00:06:16: Maybe instead of building hospitals everywhere, we should empower local staff.

00:06:20: Exactly.

00:06:21: The idea is to give nurses, maybe midwives, AI-guided tools so they can capture diagnostic quality images using stuff they might already have, like a phone.

00:06:30: Yeah.

00:06:30: It can really democratize access to specialist care and shrink those geographical caps.

00:06:35: Essentially turns the patient's own device into a potential diagnostic front-end.

00:06:40: But okay, if we're pushing diagnostic imaging onto consumer devices, what about the huge data security questions?

00:06:46: High pay compliance.

00:06:48: That must be a massive hurdle when you take this stuff outside the secure hospital network.

00:06:52: That is precisely the strategic knot you have to untangle.

00:06:55: Often the answer is a hybrid model.

00:06:57: You capture the image locally, maybe on a phone, but the complex processing happens securely in the cloud on platforms that are built with governance in mind, which actually leads perfectly into our next theme.

00:07:07: Theme three is all about governance, trust, and how AI itself is changing.

00:07:11: All this amazing speed and complexity we've talked about, quantum imaging, AI and the ICU, it all needs a really solid, believable foundation of trust.

00:07:19: And trust really seems to be the key sticking point for getting this stuff widely adopted, even while everyone knows about AI now.

00:07:26: Simon Philip Ross highlighted some crucial findings on public perception.

00:07:31: What was the main sort of tension there?

00:07:33: Well, the awareness of AI is definitely high, but public trust, it's shaky, especially the data showed among older adults and women, and that's a global trend.

00:07:43: Ross's point was strategic.

00:07:45: He basically said, trust is now a clinical feature.

00:07:48: Huh,

00:07:49: a clinical feature?

00:07:50: Yeah,

00:07:51: meaning if patients and doctors are hesitant, Adoption just stalls.

00:07:55: Doesn't matter how good the AI is.

00:07:57: So things like transparency, watching for bias, having clear audit trails they aren't nice to have, there are essential requirements now.

00:08:03: And secret birds.

00:08:03: van Roysen backed this up, laying out those specific ethical risks.

00:08:07: bias in the algorithms, data privacy, safety, who's liable if something goes wrong.

00:08:12: The consensus on prevention seems pretty clear.

00:08:14: You need strong governance rules, really rigorous testing and validation, and a big push to educate both clinicians and patients about AI.

00:08:21: And speaking of rigor, the AI itself is getting, well, smarter, more capable.

00:08:26: Jan Bigger pointed out that AI tools are quickly evolving into something called agentic AI.

00:08:32: Now, for those not familiar, agentic AI refers to systems that are more autonomous, more adaptive.

00:08:37: They don't just process data.

00:08:39: They can actually reason, plan, and act with less direct human input.

00:08:44: Think of them like semi-independent assistants.

00:08:47: So systems with agency, basically, how are they actually being used in clinics now?

00:08:51: Well, they're starting to combine different types of data, text notes, images, vital signs to help with complex stuff like suggesting differential diagnoses.

00:08:59: We're also seeing them begin to assist in robotic surgery, maybe by translating a surgeon's spoken commands into precise movements.

00:09:06: The point is, the AI is doing more than just summarizing.

00:09:09: It's starting to actively participate or intervene.

00:09:12: Okay, that definitely makes the underlying data infrastructure even more critical, especially somewhere like Europe, with its strict data sovereignty rules.

00:09:19: Dr.

00:09:20: Christian Lofert shared news about Deutsche Telekom actually acquiring a company, Synedra ITGMBH.

00:09:26: What's the strategy behind that move?

00:09:27: The goal is explicitly to build a sovereign interoperable platform for medical data across Europe.

00:09:34: This is foundational work.

00:09:36: As this agentic AI gets more powerful and acts more independently, you absolutely need secure, meaningful and crucially trustworthy ways for data to flow between different systems and countries.

00:09:48: It's a big investment saying control and trust over the data itself have to come first before you deploy the really advanced AI.

00:09:55: Right?

00:09:55: Okay.

00:09:56: Let's shift gears slightly for our last theme.

00:09:58: Theme four.

00:09:59: Remote monitoring, virtual care, and consumer empowerment.

00:10:02: How care is successfully moving beyond the hospital or clinic walls?

00:10:07: What's the current state of virtual care?

00:10:08: Is it maturing?

00:10:10: It definitely seems solidified as a core part of healthcare delivery now.

00:10:13: Teledoc Health got named Outstanding IntelliHealth by Time Magazine.

00:10:16: Chuck DeVita highlighted that.

00:10:18: It's not exactly breaking news anymore, but it does signal mainstream acceptance.

00:10:22: But it feels like the market's getting more sophisticated than just simple video calls, right?

00:10:26: Grant Kohler mentioned Teladoc is now using a tool to help hospitals actually assess their whole virtual care tech stack to find gaps maybe where they have duplicate systems.

00:10:36: That sounds like organizations are moving past just reacting and deploying towards strategically optimizing what they've got.

00:10:43: Absolutely.

00:10:44: They need to figure out what tools they actually have and how they all fit together or don't.

00:10:49: And at the same time, we're seeing patient safety almost being crowdsourced through wearables and continuous monitoring.

00:10:55: Mark D. E. Martini pointed this out.

00:10:57: Devices like those from Corsano Health let patients and their caregivers flag concerns immediately.

00:11:02: Ah, so it shifts some of that continuous monitoring burden away from staff and onto the device, empowering the patient.

00:11:09: Precisely.

00:11:09: It could significantly boost safety because you get proactive intervention potentially anywhere at home, at work, not just when the patient is physically inside the hospital.

00:11:19: So if tech is empowering consumers like this, where does it ultimately lead for the informed, connected patient?

00:11:26: Verily, Alphabet's Life Sciences Company, they made... Quite a big announcement recently aimed right at that end user.

00:11:32: Yeah, Verily launched a major sort of three prong strategic initiative.

00:11:36: Posts from Steven G, Tyler Trug and Eric Lack covered this.

00:11:40: First, there's Verily ME.

00:11:42: That's the app the consumer uses the front end.

00:11:45: It gives personalized health recommendations based on pulling together the patient's records.

00:11:49: It even has an AI companion called Violet.

00:11:51: Okay, so a consumer app tied to a personal health profile.

00:11:54: What were the other parts?

00:11:56: The second piece is the lifelong health study.

00:11:58: Basically, a huge research registry to feed data into their AI models.

00:12:03: And third, they launched Verily Pre.

00:12:05: Now, this is the critical internal piece.

00:12:06: It's their AI native precision health platform.

00:12:10: Think of Pre as the engine room, the secure governed back end designed to power all their different solutions.

00:12:16: The whole thing is a very clear strategic play towards predictive, preventive, personalized care delivered straight to the consumer.

00:12:24: And notably, they emphasized building that governance platform first.

00:12:28: So if we pull these four themes together, what's the big picture from this deep dive?

00:12:32: Well, we've definitely seen a clear industry shift towards tangible integration.

00:12:36: It's about making things work now.

00:12:37: AI giving nurses back time, quantum speeding up scans, and this huge effort really to rebuild the architecture, whether it's Europe's sovereign data push or Verily's pre-platform to create the public trust needed for this stuff to really scale.

00:12:50: If you enjoyed this deep dive, new episodes drop every two weeks.

00:12:53: Also check out our other editions on ICT and tech insights, defense tech, cloud, digital products and services, artificial intelligence and sustainability in green ICT.

00:13:02: And a final provocative thought for you.

00:13:04: We see huge companies like Oracle embedding advanced AI, maybe even GPT-V level tech across all their applications.

00:13:11: Agenic AI is starting to pop up everywhere from ICUs to genomics research.

00:13:17: So the question is, how quickly can that essential foundational work the governance, the AI literacy that people like Simon Philip Rost flagged as critical needs.

00:13:26: How quickly can that realistically catch up with the sheer speed of deployment we're seeing right now across global health care?

00:13:32: Thank you for joining us for this deep dive.

00:13:33: Don't forget to subscribe.