Best of LinkedIn: Health Tech CW 34/ 35

00:00:00: provided by Thomas Allgaier and Frenis, based on the most relevant posts on LinkedIn about health tech in CW-three-four-thirty-five.

00:00:07: Frenis is a B to B market research company working with enterprises to optimize their campaigns with account and executive insights far beyond AI.

00:00:16: Welcome, deep divers.

00:00:18: Today, we're plunging into the latest currents in health tech, pulling insights directly from the most active conversations on LinkedIn during calendar weeks, thirty-four and thirty-five.

00:00:27: That's

00:00:27: right.

00:00:28: And what really jumps out, I think, is this clear shift we're seeing.

00:00:31: Yeah.

00:00:32: What's that?

00:00:32: Well, AI seems to be moving past that initial, you know, hype cycle.

00:00:36: Ah,

00:00:36: okay.

00:00:37: Settling down a bit.

00:00:38: Exactly.

00:00:39: It's becoming more about disciplined real-world deployment across the whole healthcare ecosystem.

00:00:44: Less buzz, more building.

00:00:45: Which

00:00:46: means, I suppose, a bigger focus on tangible evidence.

00:00:48: Precisely.

00:00:49: Evidence, evolving regulations, the impact on the workforce, a huge one, and strategic alliances too.

00:00:55: So we're basically exploring how these pieces are shaping a more mature health tech landscape, moving beyond just excitement to actual implementation.

00:01:03: that affects you, our listeners navigating this space.

00:01:06: Couldn't have said it better.

00:01:07: All right, let's unpack this then.

00:01:09: Our first major theme today seems to be revolving around AI strategy, the regulatory side of things, and that really crucial element.

00:01:20: Trust.

00:01:21: Fundational stuff.

00:01:22: Totally.

00:01:22: Basic question, right?

00:01:23: As AI gets deeper into patient care, how do we make sure it's built and deployed not just effectively, but like truly responsibly?

00:01:33: That's the core challenge.

00:01:34: And we saw some really compelling takes.

00:01:36: Matt Rosenberg had this powerful argument.

00:01:39: He thinks AI should basically give clinicians

00:01:41: superpowers.

00:01:42: Superpowers?

00:01:43: Like what?

00:01:43: Things like summarizing patient history instantly, flagging concerns in real time, even auto-generating visit notes.

00:01:50: Ah, freeing up their time for actual patient care.

00:01:52: Exactly.

00:01:53: And Rafael Satiro... kind of doubles down on this, stressing this customer-centric product mindset.

00:01:58: Meaning?

00:01:58: Meaning you have to deeply understand the need and the problem before you even think about the AI solution or the algorithm.

00:02:05: Right.

00:02:05: Otherwise, you're just building tech for tech's sake.

00:02:07: You got it.

00:02:07: Building a solution, looking for a problem.

00:02:10: That's a vital point.

00:02:11: But, you know, not everyone's jumping on the AI bandwagon immediately.

00:02:15: Reza Zahiri had an interesting counterpoint.

00:02:17: Yeah.

00:02:18: He suggested that for a company like Philips, the priority should maybe be Focus, not AI.

00:02:24: Interesting take.

00:02:25: Why?

00:02:26: He argues chasing the latest AI trend can distract from core internal stuff, operational inefficiencies, maybe not getting clear ROI on R&D.

00:02:36: So

00:02:36: fix the fundamentals first.

00:02:37: Kind of.

00:02:38: Simplify operations, sharpen R&D on core strengths like medical imaging, patient monitoring.

00:02:44: It's a reminder that the flashiest solution isn't always the smartest move right now.

00:02:48: That tension definitely came through.

00:02:50: between the cutting edge and the foundational stability.

00:02:54: Myeongchaw, for example, he speculated that AI, not traditional platforms like Epic, AI might become the real aggregator in healthcare.

00:03:02: Wow.

00:03:02: Okay.

00:03:03: How so?

00:03:03: Through

00:03:04: AI agents, maybe new interoperability rails like Tefka, you know, the trusted exchange framework, potentially changing the whole game.

00:03:11: But there's always a but, isn't there?

00:03:13: Well, yeah.

00:03:14: Simon Philip-Brosch highlighted the situation in Europe.

00:03:16: They have the tech, but deployment is slow.

00:03:18: Why is that?

00:03:19: fragmented data, outdated IT infrastructure, and just general hesitancy.

00:03:24: Worries about trust, liability, regulatory clarity.

00:03:28: It's complex.

00:03:29: The vision's there, but the practical hurdles.

00:03:33: Significant.

00:03:33: And that brings us right into the legal and ethical minefield.

00:03:37: Jacqueline Batchford, she's an in-house lawyer in cancer care, described it as navigating an incomplete AI regulatory landscape.

00:03:44: Incomplete.

00:03:45: Yeah, that sounds about right.

00:03:46: She pointed out the EU AI Act classifies healthcare AI as high risk.

00:03:51: That means strict transparency, human oversight.

00:03:53: Heavy requirements.

00:03:54: Definitely.

00:03:55: And at the same time, you've got GDPR protecting sensitive data, which is vital, but it can make sharing data for diverse AI training data sets tricky.

00:04:03: A real balancing act.

00:04:05: She did mention the European health data space could be a game changer though.

00:04:08: potentially offering a framework for that balance.

00:04:10: Building on that, Deepak Pant highlighted that AI and medical devices, it's mostly seen as software as a medical device or SAMD.

00:04:19: And the EU AI Act is setting a pretty high bar for requirements there.

00:04:24: Then Michelle, Brazil raised some really critical questions.

00:04:26: Like, what will the regulatory framework actually look

00:04:29: like?

00:04:29: Good question.

00:04:30: What needs defining?

00:04:32: Things like clinical utility for these adaptive algorithms.

00:04:36: Who's accountable?

00:04:36: How do we get real world evidence at scale to build that trust?

00:04:41: Without clarity, AI innovation could stall.

00:04:44: And trust.

00:04:45: It just keeps coming back, doesn't it?

00:04:47: It feels like the absolute cornerstone.

00:04:49: It really is.

00:04:50: Fonique Marquet made a strong case for developing inclusive

00:04:53: AI.

00:04:54: Meaning training on diverse data sets.

00:04:56: Exactly.

00:04:57: To ensure fair, accurate, and inclusive care across all demographics, Khalil Ramon echoed this perfectly.

00:05:03: He just said, it all comes down to trust in health care.

00:05:05: Simple,

00:05:05: but true.

00:05:06: Build on patient safety, data security, system reliability, especially now with AI transforming testing and diagnostics.

00:05:13: That's a powerful goal.

00:05:14: But Rashmi Raghavendra brought a bit of a reality check.

00:05:18: Oh, yeah.

00:05:18: What was her take?

00:05:20: She observed that AI in health care might be hitting a sort of pause point.

00:05:23: The hype, she thinks, is settling.

00:05:25: And the hard work is starting.

00:05:27: That's it.

00:05:28: The hard work is starting to get sustained real impact.

00:05:31: She even compared it to the early twenty tens, lots of big ideas, but making them stick was tough.

00:05:38: It feels like a moment for grounded effort.

00:05:41: That's a really good way to put it.

00:05:42: OK, so.

00:05:44: Moving from those big picture strategies and challenges, let's get into some specific examples where AI is already making a difference in actual clinical workflows.

00:05:53: Yeah, let's look at where the hard work is paying off.

00:05:56: Where are we seeing tangible value right now?

00:05:58: Well, Yahya M highlighted a few key areas.

00:06:01: Medical imaging and diagnostics, that's a big one.

00:06:04: Right.

00:06:04: Clinical documentation automation.

00:06:07: which is huge for tackling clinician burnout.

00:06:10: Definitely hear a lot about that.

00:06:11: And risk stratification using predictive alerts.

00:06:14: These aren't just future promises.

00:06:15: AI apps are already driving real value here, sometimes matching or even beating human experts.

00:06:21: That's genuinely impressive.

00:06:23: And I saw a really compelling study from Louis Attala.

00:06:25: Oh, the ICU discharge one.

00:06:27: Yeah, a deep learning model predicting forty-eight hour post ICU discharge mortality.

00:06:33: It hit an AUC of point nine three which

00:06:35: is very very good.

00:06:37: accuracy

00:06:38: Exactly.

00:06:39: and the amazing part it only uses six simple features age basic vital signs.

00:06:45: So it prioritizes practicality with a simple design right

00:06:49: makes it easier to integrate easier for clinicians to trust.

00:06:52: Sometimes simplicity really is key.

00:06:54: and speaking of clinician trust and usability Craig Joseph from Stanford healthcare shared insights from their pilot using generative AI.

00:07:02: What were they using it for?

00:07:03: Drafting patient-friendly explanations for lab results, imaging, pathology reports, stuff like that.

00:07:10: Aimed

00:07:10: at reducing burnout and helping patients understand better.

00:07:13: Exactly.

00:07:14: Clinicians found it usable and useful.

00:07:17: Still need some refinement for clinical nuance, obviously, but the potential is there.

00:07:22: It's about assisting, not replacing.

00:07:23: Good

00:07:24: point.

00:07:24: And Andrea Petraca highlighted something very specific to dental CT with tin filter technology.

00:07:29: Ah, yes, in human and vet medicine.

00:07:31: Yeah, enhancing diagnostics and treatment planning, especially in the Maxwell Facial area, shows AI's precision even in quite specialized fields.

00:07:38: Okay, so we've seen specific applications.

00:07:41: Now let's shift focus a bit to the foundations.

00:07:44: The digital infrastructure, the interoperability.

00:07:47: How systems talk to each other.

00:07:49: critically important because without that solid base even the smartest AI is limited, isn't it?

00:07:54: Absolutely

00:07:55: feels that way.

00:07:55: Yeah, Robert Wagner shared this insight.

00:07:57: apparently healthcare execs and clinicians universally rank interoperability as the hashtag one priority for digital transformation number

00:08:06: one.

00:08:07: Wow

00:08:07: his argument.

00:08:08: without breaking down those data silos AI just can't deliver its full potential period

00:08:13: makes total sense and Seema verma from Oracle seemed to echo that commitment.

00:08:18: Yes, pledging support for the CMS Digital Health Ecosystem, aiming for a secure, interoperable, standards-based, and AI-enabled system all focused on better outcomes.

00:08:28: That sounds like a huge undertaking, but necessary.

00:08:31: Mary Beth Siegerlin, also from Oracle, pointed to a more immediate win, though.

00:08:35: Oh, what was that?

00:08:36: AI streamlining prior authorization.

00:08:38: Ah.

00:08:39: The bane of many clinicians' existence.

00:08:41: Exactly.

00:08:42: Cutting red tape, reducing delays, letting clinicians focus on care.

00:08:46: She mentioned an industry pledge, aiming to answer eighty percent of electronic prior auth requests in real time by twenty twenty seven.

00:08:54: Eighty

00:08:55: percent in real time.

00:08:56: That would be massive.

00:08:57: Yeah.

00:08:57: Driven by AI, machine learning, RPA.

00:09:00: Yeah.

00:09:00: Significant shift if they hit it.

00:09:02: And beyond just prior Roth, modernizing the core EMR and hospital ops platforms is still crucial, right?

00:09:09: For basic efficiency.

00:09:10: For sure.

00:09:11: Jessica Matheson showcased platforms like care.ai, smart connected care.

00:09:16: What does that involve?

00:09:17: Things like virtual nursing, ambient monitoring.

00:09:19: Yeah.

00:09:20: Basically helping care teams enhance patient care and boost their own efficiency.

00:09:24: Smart.

00:09:24: And Karthik Sukumar detailed how AI agents in Oracle Cloud infrastructure are already making waves.

00:09:30: Not just in health care.

00:09:31: No, across industries, but definitely including health care.

00:09:33: Yeah.

00:09:33: Integrating LLMs, ROG, vector search.

00:09:36: It's

00:09:36: about making enterprise operations much more intelligent.

00:09:39: It feels like product strategy is shifting to match that need for scalability.

00:09:44: Amit Joshapura mentioned Verily's pivot.

00:09:46: Towards scalable, sauce-driven impact in medtech, yeah.

00:09:48: Yeah.

00:09:49: With platforms like LightPath for chronic care, Workbench for secure biomedical data.

00:09:55: moving towards broader, more accessible tech solutions.

00:09:58: And it's not just the US.

00:10:00: Tomás Domino described the UK making some bold moves.

00:10:03: Like what?

00:10:04: Expanding the NHS app into a proper digital front door, scaling up total triage models.

00:10:11: AI-powered triage.

00:10:12: Yeah, expected to unlock millions of GP appointments annually, providing workflow support for staff, guided self-navigation for patients.

00:10:20: Potentially a huge efficiency gain over there.

00:10:23: Okay, let's move from the digital infrastructure to the well physical side the tangible innovations medtech and robotics.

00:10:30: What's happening there?

00:10:31: Surgical

00:10:32: robotics is still a really hot area.

00:10:33: Just an AI from Medtronic sees their Hugo system as a key growth driver.

00:10:37: Especially

00:10:37: with US FDA approval pending, they already have the CE mark in Europe, right?

00:10:41: Correct.

00:10:42: And Christian Mazzi celebrated a milestone in robotic heart surgery using Percival technology, really highlighted how collaboration is advancing options for patients.

00:10:50: That's great to hear.

00:10:51: And the evidence back in these procedures is growing too.

00:10:54: Definitely.

00:10:54: Dr.

00:10:55: Quinland D. Buchlach's study in general surgery showed better efficiency, better safety, no intraoperative complications.

00:11:01: And a quick learning curve.

00:11:03: Yeah, median docking time dropped significantly after just ten procedures.

00:11:07: Plus, Frank R. Busley highlighted discussion around tristapal in robotic bariatric surgery, focusing on reliability, secure outcomes.

00:11:15: So the tech is proving itself out in specific procedures.

00:11:18: It is.

00:11:19: And innovation isn't just about surgery.

00:11:21: Tiraj Jindal introduced something different.

00:11:24: Fluid filled glasses?

00:11:26: Fluid filled glasses.

00:11:27: For what?

00:11:28: To combat motion sickness.

00:11:30: They use dynamic fluid motion to give your brain a more complete visual picture, matching what your inner ear is sensing.

00:11:36: Huh.

00:11:37: Clever trick to fool the brain.

00:11:38: Pretty neat solution.

00:11:39: And Deepak Pant specifically called out diagnostics and imaging radiology AI plus monitoring and wearables.

00:11:45: Like smartwatches spotting irregularities.

00:11:47: Exactly.

00:11:48: And surgical support like robotic precision.

00:11:50: Those are key segments where AI is really being touched in medical devices, driving accuracy and automation.

00:11:56: All right, let's shift gears again.

00:11:58: How is AI changing the game in pharma, R&D, and this idea of precision medicine?

00:12:05: This feels like we're some truly transformative potential

00:12:08: lies.

00:12:09: Absolutely.

00:12:10: Duncan Porter's post on... AI revolutionizing antibiotic discovery was seriously exciting.

00:12:15: How

00:12:16: so?

00:12:16: He

00:12:16: detailed breakthroughs like finding Hallison an antibiotic candidate in just three days using AI.

00:12:22: Three days.

00:12:23: And discovering a Boston against drug-resistant bacteria.

00:12:26: It just shows AI's power to smash timelines and find novel approaches, crucial for fighting superbugs.

00:12:32: That's

00:12:32: incredible speed.

00:12:34: And what about precision medicine?

00:12:35: Michael Revhan was talking about its evolution.

00:12:37: Yeah, he explored the dimensions of precision.

00:12:39: Where does more accuracy matter.

00:12:41: Temporal precision timing interventions just right.

00:12:44: Spatial precision targeting treatments accurately.

00:12:47: And endotype precision getting down to molecular individual level.

00:12:51: And digital twins are helping with that complex learning.

00:12:53: Increasingly

00:12:54: so, yes.

00:12:55: Facilitating that kind of tailored approach.

00:12:57: We also saw a specific R&D win from Machado Valapor.

00:13:00: Right, in bioprocessing.

00:13:01: Significant productivity gains using autonomous control software in CHO perfusion.

00:13:08: What kind

00:13:08: of gains?

00:13:09: Around a seventy percent jump in volumetric productivity, stable operation, high viability, measurable real-world improvement in efficiency.

00:13:18: And Wolfgang Schleifer also saw AI's promise here.

00:13:21: Yeah, he observed it promises to recode the difficult process of drug discovery, streamline clinical trials, revolutionize areas like cardiology.

00:13:30: But he acknowledged challenges too.

00:13:32: Oh yeah,

00:13:32: high trial costs, regulatory lag.

00:13:35: Those are still big hurdles, but the potential of AI to speed things up and refine them is undeniable.

00:13:41: And Verily seems active here too.

00:13:43: Eric Yang shared research.

00:13:44: Yes, published in JMIR.

00:13:46: an LLM-powered workflow helping with sustained behavioral change in chronic disease management.

00:13:51: How does that

00:13:51: work?

00:13:52: It identifies individual barriers, like why someone isn't sticking to a nutrition plan with over ninety percent accuracy.

00:13:58: Then it delivers personalized tactics, behavioral science nudges.

00:14:00: Using AI to really personalize coaching.

00:14:03: Impressive.

00:14:03: Very

00:14:03: much so.

00:14:04: Meeting patients where they are.

00:14:05: Okay, let's pivot now to the people.

00:14:07: The workforce.

00:14:08: How is all this health tech impacting clinicians, support staff, daily operations.

00:14:14: This is crucial.

00:14:15: Jason Case expressed a strong belief that responsible adoption of AI will help close critical gaps.

00:14:22: Addressing worker shortages.

00:14:23: Exactly.

00:14:24: While empowering clinicians to focus on patients, Robert Slepen reinforced this, citing Harvard professor Christopher Stanton.

00:14:31: What is

00:14:31: Stanton's point?

00:14:32: That AI should enhance the workforce, not displace jobs.

00:14:35: His example was radiologists, they're as busy as ever, but using AI tools lets them do more.

00:14:42: efficiently.

00:14:43: Augmentation, not replacement.

00:14:45: That's

00:14:45: the ideal scenario, isn't it?

00:14:46: And Bird Van Meers from Phil's talked about human-centered health tech.

00:14:50: Yeah, how AI can alleviate stress by unifying data, cutting workloads, simplifying admin tasks, basically giving clinicians back time and mental energy.

00:14:58: You mentioned a stat too.

00:14:59: Right.

00:14:59: Seventy-four percent of healthcare pros believe AI can improve patient access.

00:15:04: And Meenal Shah emphasized human-centered design, again reducing the cognitive tax from constant interruptions, streamlining workflows, cited the Epic Monitor as a good example.

00:15:13: Makes sense.

00:15:13: Yeah.

00:15:14: But there was a counter Wasn't there something a bit concerning?

00:15:17: Ah, yes.

00:15:19: Rudolph Wagner's post.

00:15:21: This was fascinating and maybe a bit worrying.

00:15:23: What did it say?

00:15:24: He highlighted a study in Lancet gastroenterology and hepatology.

00:15:29: Doctors using an AI tool for colonoscopies after just three months were significantly worse at finding pre-cancerous growths on their own.

00:15:37: Worse, really?

00:15:38: Yeah, it raises serious concerns about de-skilling.

00:15:41: The idea that relying too much on tech might erode fundamental human skills.

00:15:46: Wow, that de-skilling idea is... like rolling too much on GPS and forgetting how to navigate.

00:15:53: But the stakes here are infinitely higher.

00:15:55: Exactly.

00:15:56: It's a stark reminder.

00:15:56: we need to manage that balance carefully.

00:15:59: Which makes it important that companies like Oracle, as Tamer Taba highlighted, are getting recognized for workforce management tools that prioritize a human-oriented approach.

00:16:07: Not just tech for tech, Sam.

00:16:08: Right.

00:16:09: And speaking of the human element, we can't ignore diversity.

00:16:12: Jewariah Khan made a strong case for increased gender diversity in med tech leadership.

00:16:17: Why is that so important?

00:16:19: Stronger teams, better customer connections, ultimately better healthcare outcomes.

00:16:23: She's stressed, talent, not gender, defines every boardroom conversation.

00:16:28: Good point.

00:16:29: And Vicki Britton's analysis touched on this too.

00:16:32: regarding women's health AI.

00:16:34: Yeah, patient perceptions of AI and menopause care.

00:16:37: Positive uses for advocacy, yes, but also growing concern.

00:16:40: A lack of women developing the AI could lead to bias against women's conditions.

00:16:45: So inclusive design needs inclusive developers.

00:16:47: Absolutely, at every single level.

00:16:49: Okay, let's wrap up our main themes with something absolutely critical.

00:16:53: Cyber security and data integrity.

00:16:56: As AI expands, this has to be rock solid.

00:16:59: Non-negotiable.

00:17:00: LT-CDR Amit Pal Singh's upcoming talk at cyber.ai summit, really underscores this.

00:17:07: He's focused on securing critical infrastructure in an AI-driven world.

00:17:11: Building

00:17:11: secure resilience systems for healthcare tech, vital.

00:17:14: Given the sensitivity of health data, couldn't be more important.

00:17:17: And Mark Rakhmalovic talked about building trust in the data itself, using blockchain.

00:17:22: Yeah, exploring that intersection, covering things like federated learning.

00:17:25: Where

00:17:25: you share model insights, not raw data.

00:17:27: Exactly.

00:17:28: Plus, authenticated provenance and supply chains think pharmaceuticals and ensuring high compliance throughout, using tech to build verifiable trust.

00:17:36: And we're seeing collaboration on this front too.

00:17:39: Pretty Panda shared news about the Cancer AI Alliance.

00:17:42: Right.

00:17:43: Four major NCI-designated cancer centers funded by tech giants like AWS and Microsoft, they're joining forces.

00:17:50: To do what exactly?

00:17:51: Revolutionize cancer research using multimodal data and federated learning.

00:17:55: So they collaborate securely, compliantly, without exposing raw patient data.

00:18:00: It's a massive step for privacy-preserving research.

00:18:03: That's huge.

00:18:04: And Cal O'Roma's point about testing comes back here too.

00:18:06: Definitely.

00:18:07: Robust testing.

00:18:08: is crucial to protect patients, ensuring flawless diagnoses, treatment plans, data integrity.

00:18:13: It's paramount in this AI revolution.

00:18:15: But there's still a regulatory challenge, isn't there?

00:18:17: Jan Beger raised a point.

00:18:19: A big one.

00:18:20: He argues current medical device regulations are basically inadequate for advanced AI agents.

00:18:25: Why

00:18:25: inadequate?

00:18:26: Because of their autonomy, their adaptability, their broad functionality.

00:18:30: It doesn't fit the old models, his implication.

00:18:33: Without bold regulatory evolution, Truly autonomous AI agents in clinical care might just stay out of reach.

00:18:41: So regulation needs to evolve as fast as the tech.

00:18:44: or risk holding it back.

00:18:45: That's

00:18:45: the challenge, a really urgent one.

00:18:47: Okay, wow.

00:18:48: Let's try and unpack all of that.

00:18:49: We've covered a lot of ground.

00:18:51: We've seen health tech really poised for incredible change.

00:18:54: AI giving clinicians those superpowers, transforming drug discovery.

00:18:58: But it all comes back to fundamentals, right?

00:19:00: Trust, careful regulation, human-centered design.

00:19:04: Absolutely.

00:19:04: And we even touched on that.

00:19:07: intriguing, maybe slightly unsettling idea of de-skilling.

00:19:10: Yeah,

00:19:11: that's one to watch.

00:19:12: So the big question for you listening, the professional navigating this space, as AI and advanced tech keep evolving at this incredible pace, how do you see the ultimate balance shaping up?

00:19:23: That balance between pushing technological boundaries and preserving that uniquely human element, the empathy, the expertise, it's essential for the best possible patient care.

00:19:31: It's definitely something we all need to be thinking about

00:19:33: actively.

00:19:34: For sure.

00:19:35: Well, if you enjoy this deep dive, remember, new episodes drop every two weeks.

00:19:39: You can also check out our other editions on ICT and Tech Insights, DefenseTech, Cloud, Digital Products and Services, Artificial Intelligence, and Sustainability in Green ICT.

00:19:50: Lots

00:19:50: to explore.

00:19:51: Definitely.

00:19:51: Thanks so much for joining us today.

00:19:53: Make sure to subscribe and join our community to stay well informed.