Best of LinkedIn: Health Tech CW 10/ 11

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

00:00:08: Frennis equips product and strategy teams with market and competitive intelligence to navigate

00:00:18: glowing holograms, doctors and virtual reality headsets.

00:00:22: Maybe artificial intelligence making these dramatic life-saving diagnoses in the blink of an eye?

00:00:28: Right!

00:00:28: The whole sci-fi movie set

00:00:30: up Exactly.

00:00:30: but according to the brightest minds in health tech right now...the future of healthcare shouldn't be dramatic at all.

00:00:36: In fact it actually needs to be incredibly predictably boring.

00:00:40: Boring

00:00:41: I love that.

00:00:41: Yeah, welcome to The Deep Dive!

00:00:43: We are here to help you make sense of the ground shifting beneath your feet in this industry.

00:00:47: And man what a shift it is.

00:00:49: So for today's deep dive we're dissecting the top health tech trends That really dominated professional conversations across LinkedIn In calendar weeks ten and eleven of twenty-twenty six.

00:00:59: yeah there was alot of chatter

00:01:00: A ton...and the overarching mission For You Today Is exploring This massive Industry Wide Pivot.

00:01:06: i mean We Are Officially Leaving The Era Of Tech Experimentation.

00:01:10: Right, that phase where hospitals were basically just throwing shiny new tools at the wall to see what sticks.

00:01:15: Exactly!

00:01:16: That phase is over.

00:01:17: now The mandate is all about organizational readiness.

00:01:20: right and measurable clinical impact.

00:01:23: Yeah And to really unpack that shift we are going to navigate through three specific deeply connected clusters.

00:01:29: today We're looking how AI workflow integration is actually functioning on the ground, then the evolution of healthcare data platforms into like foundational infrastructure and finally How all this digital maturity is scaling physical robotics in virtual care.

00:01:44: Awesome well Let's jump right into the deep end with the biggest buzzword in the room Right?

00:01:48: Oh always a I.

00:01:50: but look we are setting aside The hypothetical futuristic models here.

00:01:55: We want to know what happens when these algorithms Actually hit the hospital floor And to set the stage.

00:02:01: Taha Cashout posted this fascinating argument recently.

00:02:06: He made the case that the future of healthcare AI needs to be, well exactly what you said in the intro.

00:02:11: Completely

00:02:12: boring?

00:02:12: Completely boring!

00:02:13: Yeah I mean it sounds counterintuitive but he argues its really only way forward.

00:02:19: health care won't be transformed by these massive stand-alone AI models that act like a digital oracle giving you answers right.

00:02:26: The transformation comes from Think of a control plane Kind of like an air traffic control system that just sits invisibly behind the scenes.

00:02:36: Okay,

00:02:36: so it's not front and center

00:02:37: exactly.

00:02:38: It's a system of trusted automated workflows coordinated entirely around the patient's journey.

00:02:43: So it's NOT about the software you know showing off its predictive power?

00:02:46: It's about the total absence of friction.

00:02:49: Right because I mean friction in a clinical setting doesn't just cost money but costs lives.

00:02:54: Vivica Rora actually posted a fantastic five-layer framework, explaining the mechanics of how this integration is happening inside existing electronic medical records or EMRs.

00:03:06: He pointed out that AI in healthcare has finally moved past that brute force phase...

00:03:11: The phase of just scraping unstructured text?

00:03:14: Yeah exactly!

00:03:14: Just scraping PDF's from raw data lakes That old news.

00:03:19: We are reaching what he defines as layer five.

00:03:21: Oh Layer Five the clinical AI agents.

00:03:24: You got it, and the distinction here is super mechanical.

00:03:27: instead of a doctor opening a separate browser tab logging in asking chatbot to analyze patient's history which

00:03:34: nobody has time for right?

00:03:35: Instead all that The AI agent is embedded directly into native UI EMR.

00:03:40: so drafting notes summarizing charts teeing up orders all just background defining characteristic.

00:03:46: layer five does not force clinicians change their physical habits.

00:03:50: no extra clicks

00:03:51: Zero extra clicks.

00:03:52: Well, okay I have to push back a little bit on this idea of like seamless integration though.

00:03:56: Okay

00:03:57: late on me

00:03:57: because Seamless assumes the hospital runs exactly The way the manual says it does which as we know is rarely how things actually play out On the

00:04:06: floor.

00:04:06: Oh that was very true.

00:04:07: Yeah Meenal Shah wrote A brilliant post That serves As massive reality check here.

00:04:13: She pointed Out technology often fails in health care Because developers completely ignore the Susan layer.

00:04:20: The Susan layer that is honestly the best description of hospital operations I've heard all year.

00:04:26: Isn't it great?

00:04:27: It's so relatable for anyone who has worked in complex operations, i mean...the Susan layer is that invisible operational network Of nurses technicians clinicians Who actually keep clinics running through a million undocumented work around.

00:04:40: Right, because the documented standard operating procedure might say one thing but The Real Playbook

00:04:45: is what resides exclusively inside Susan's brain.

00:04:48: Exactly!

00:04:49: Susan knows that printer on third floor jams Tuesday so she manually routes lab requests to second floor.

00:04:55: So if your shiny new layer five AI agent assumes the documented workflow Is the real workflow it will break immediately.

00:05:03: If it tries to pave over the Susan layer without understanding the underlying mechanics of why Susan does what she does, clinicians will literally just build a new workaround to avoid the new technology.

00:05:13: Yeah

00:05:13: they'll just ignore entirely.

00:05:15: and that tension is exactly why organizational governance Is currently lagging so far behind the underlying technology.

00:05:22: You can't just deploy an algorithm into chaos.

00:05:24: No you really cant.

00:05:25: Bill Russell highlighted this divide perfectly.

00:05:28: He drew a very sharp line between administrative deployment and clinical deployment.

00:05:33: on the administrative side, scaling is incredibly fast like he pointed to Intermountain Health.

00:05:37: Oh yeah didn't they save a massive amount of time?

00:05:40: Massive!

00:05:40: They saved twenty thousand IT hours And resolve thirty five thousand HR increase just using back-end AI agents?

00:05:47: Well

00:05:47: sure because the stakes of an HR bot making mistake are fundamentally lower than a Clinical Bot Making A Mistake

00:05:53: Precisely The Point.

00:05:55: Clinical deployment requires deep systemic accountability.

00:05:58: You cannot let an agent run wild with patient care, just mapping its own logic over the Susan

00:06:04: layer.".

00:06:04: Which is why Jan Beger brought this into focus right?

00:06:07: He shared the American Medical Association's eight-step playbook for AI governance.

00:06:12: Yeah, the AMA book The

00:06:13: core message there is that health systems must build AI governance structures like interdisciplinary working groups continuous auditing loops liability frameworks before they even think about deploying the algorithms themselves?

00:06:25: The governance is the infrastructure

00:06:27: and the gap between the marketing hype of these tools in reality on the ground.

00:06:31: That was a huge theme at the recent VV twenty twenty six conference.

00:06:35: Sachin Pohar shared his observations straight from the floor, noting that terminology is vastly outpacing actual technical understanding right now.

00:06:43: Oh

00:06:43: for sure!

00:06:43: Everyone's throwing around buzzwords.

00:06:45: Exactly Every vendor is pitching ambient AI and autonomous agents but very few hospital buyers understand heavy lifting required to integrate them.

00:06:55: He made a crucial point for you The Listener If your are health tech professional Right Now You must actively upskill.

00:07:01: You have get your hands dirty Yeah...You

00:07:03: Have To Get Your Hands Dirty building and testing these tools in your daily workflows, otherwise you literally have no framework to separate the actual signal from the noise.

00:07:13: But here is the catch with empowering The Susan layer and deploying these clinical agents.

00:07:19: Susan cannot make smart invisible decisions if the patient's comprehensive history is locked into proprietary PDF at a different hospital system across town

00:07:28: of the interoperability nightmare.

00:07:29: Exactly, you cannot have seamless AI workflows without pristine borderless data.

00:07:35: which brings us to our second cluster The battle for personal health data ecosystem where healthcare data platforms are morphing into strategic infrastructure.

00:07:44: It really is the foundation everything else has built upon because if data is fragmented, AI is hallucinating and we are seeing big tech moving aggressively to solve this defragmentation problem.

00:07:55: Yeah a massive example from past couple of weeks was the launch Microsoft Co-Pilot Health.

00:08:01: Both Artie Moudallar and Mustafa Suleiman discussed mechanics at length.

00:08:07: It is basically designed to provide personalized insights by connecting your wearable device data directly with medical records.

00:08:13: But wait, what about privacy?

00:08:15: Because handing all that over to Microsoft sounds risky!

00:08:18: Well, what's most important about that launch is the architectural choice around privacy.

00:08:23: The data is encrypted locally and crucially it is not used to train Microsoft's broader AI models.

00:08:29: Oh

00:08:29: wow!

00:08:30: Okay That big deal.

00:08:31: Yeah It fundamentally changes the trust dynamic for users because you aren't feeding your sensitive health history into a giant corporate language model...

00:08:39: ...that makes alot of sense.

00:08:40: And Silliman actually noted very specific deeply human use case driving this adoption.

00:08:46: They are seeing heavy usage of this tool on mobile phones late at night by the sandwich generation.

00:08:52: The

00:08:53: sandwich generation?

00:08:54: Yeah, so these are adults who were simultaneously caring for their young children and they're aging parents at the same time.

00:08:59: Oh man

00:08:59: that is a tough spot to be in!

00:09:01: It's exhausting...they are awake at two AM trying to make sense of complex symptoms an overlapping test results.

00:09:07: So giving them localized tool to synthesize that data, it's just a game changer.

00:09:12: Microsoft is hardly alone in this space either.

00:09:15: Over at SXSW we saw the US launch of The Verily May app.

00:09:20: Nick Melrose and Farrot Rajagopal highlighted how this platform is centering around an AI companion named Violet.

00:09:26: Yeah, Violet uses guided conversations for symptom navigation and precision health And it utilizes your actual longitudinal medical history to give you context-aware support rather than just returning a generic web search.

00:09:40: Moving from generalized search To personalized medical intelligence Is powerful But look Here's where it gets really complicated and I have to ask the hard question.

00:09:48: Go for with all these different tech giants, Microsoft, Verli whoever else hoarding our personal data into their specific apps To build these personalized narratives aren't we just creating newer shinier silos?

00:10:01: That is the fear right.

00:10:02: instead

00:10:02: of your data being stuck at a local community hospital It has now trapped in a specific tech ecosystem.

00:10:09: Who actually owns this and how do we govern it to ensure interoperability?

00:10:13: That is the multi-billion dollar question.

00:10:16: And honestly, it's one that industry is terrified to fully answer because if you follow this logic into its necessary conclusion You arrive at Alexa Burke King argument which fundamentally challenges current business models of big tech.

00:10:28: She argued real world health data needs to be governed as a public utility A Public

00:10:33: Utility Like Water or Electricity.

00:10:35: Think of it exactly like the electrical grid.

00:10:37: You don't own the electricity flowing into your house, and you certainly don't care which specific power plant generated it—you just plug your lamp in a standardized socket…and turns on!

00:10:47: King is saying health data needs to work the exact same way through Federation.

00:10:51: We need to shift our mindset away from asking who's proprietary data asset is this?

00:10:56: To ask what are we collectively obligated?

00:10:59: Wow,

00:11:01: that's a huge paradigm shift.

00:11:02: It is!

00:11:03: If we continue to treat health data as fragmented proprietary asset locked in corporate silos We severely limit our ability drive systemic population level innovation.

00:11:15: And if don't fix the underlying data architecture The downstream effects on algorithms are mathematically dangerous.

00:11:22: Alicia L. Wurstuk brought up a critical caveat regarding this at the Health.Tech Global Summit in Basel.

00:11:28: She pointed out that we currently suffer from massive data.

00:11:30: gender gap.

00:11:31: Yes,

00:11:32: it is profound structural flaw.

00:11:34: and how we built predictive models?

00:11:36: Historically medical research clinical trials and general data collection have severely underrepresented female biology.

00:11:43: So Wurstik's point is purely mechanical one.

00:11:46: If we train tomorrow's cutting-edge diagnostic AI on yesterdays' biased, malecentric datasets... We aren't actually innovating.

00:11:55: We're

00:11:55: just reinforcing old biases

00:11:57: Exactly!

00:11:57: We are just taking our historical blind spots and scaling them infinitely through automation.

00:12:03: You cannot achieve true precision medicine if the foundational data lacks inclusivity.

00:12:08: It's a classic garbage in-garbage out problem, but at a societal scale.

00:12:12: And solving that gap means we have to radically expand where and how we collect data.

00:12:16: We can no longer rely solely on episodic hospital records

00:12:19: right?

00:12:20: Dr Ben Murthop who highlighted a fascinating solution to this.

00:12:23: he pointed out That his company Sarah has collected three hundred billion data points entirely from home health care visits.

00:12:29: wait

00:12:29: Three hundred billion three

00:12:31: hundred bill just

00:12:31: by monitoring people in their living rooms.

00:12:33: Yeah, by utilizing IoT devices smart sensors and routine caregiver inputs to monitor the delta in daily habits.

00:12:41: Because as he notes The most critical changes in a patient's health happen at home days or weeks before someone ends up in an emergency room.

00:12:49: That makes total sense.

00:12:50: By capturing the very small signals of health deterioration like a slight change in mobility, a skipped medication.

00:12:57: A variation and sleep patterns in the real world you build predictive models that actually intervene and prevent hospitalization entirely.

00:13:05: So capturing that continuous federated data at home is the vital first step.

00:13:10: but that leads us directly into our third and final theme Because once we have pristine data and the clinical agents are processing it in the background, We're seeing an explosion.

00:13:20: And how that digital intelligence translates into physical action?

00:13:23: The physical application right

00:13:24: specifically the massive scaling of clinical precision In the operating room alongside virtual care delivery.

00:13:31: We've officially moved from proof-of-concept to proof of economics.

00:13:34: This is where the digital control plane finally moves the physical needle.

00:13:37: yes

00:13:38: Let's look at surgical robotics.

00:13:40: Several posts from these past two weeks highlighted the rapid global adoption of Medtronic's Hugo RAS system.

00:13:47: James Porter and Rajik Kamal discussed mechanics on its unique modular design, which

00:13:51: is really cool!

00:13:52: It is imagine a traditional surgical robot right?

00:13:55: it is typically a massive immovable forklift of machine that takes up half operating room and forces the surgical team to work around

00:14:03: it.

00:14:03: Right totally inflexible

00:14:05: Exactly.

00:14:06: The Hugo System however uses independent arm carts.

00:14:09: It is more like having four independent, highly skilled mechanical assistants standing exactly where you need them.

00:14:15: This modularity gives surgeons unparalleled flexibility to access a larger area of the patient's abdomen without having to constantly stop surgery to reposition heavy equipment.

00:14:25: And it's scaling rapidly.

00:14:26: I mean, Frederick Laine and celebrated its successful first implementation at Maastricht UMC Plus in the Netherlands.

00:14:32: What is truly fascinating about The Hardware Evolution though Is how it is seamlessly merging with intelligent software?

00:14:38: The hardware has no longer just a mechanical tool.

00:14:40: It is becoming smart context-aware ecosystem.

00:14:44: Well, Christian Ballowider and Holger Steffen posted about the recent FDA clearance of the stealth XIS system for spine surgery.

00:14:52: This system intelligently integrates pre-operative planning, real time spatial navigation and robotic execution into a single adaptive platform.

00:15:02: So instead just being like...a highly precise drill.

00:15:05: it is acting as an GPS and active steering assist

00:15:10: Exactly.

00:15:11: It understands the anatomy it is navigating, and this type of intelligent automation isn't reserved exclusively for highly complex spine surgeries either.

00:15:19: Adrienne James highlighted how a system called Viewpoint Echo Pilot is utilizing AI in cardiac imaging.

00:15:24: The AI automates manual measurement steps and subsequent documentation.

00:15:31: repetitive motion injuries, just from the sheer volume of clicking and adjusting they have to on their machines all day.

00:15:37: That is the crucial takeaway.

00:15:38: The technology is drastically reducing the thousands physical clicks required per shift.

00:15:44: It's utilizing AI not only process a medical image faster but actively protect the physical wellness and ergonomic longevity of the clinician.

00:15:53: Alright.

00:15:53: so let look at this board.

00:15:54: here The evidence is overwhelmingly clear.

00:15:56: The advanced robotic tech works and improves access.

00:16:00: The AI reduces physical injuries for staff.

00:16:04: And on the virtual care side, Simcon pointed out that hospital at home models are currently delivering thirty to fifty percent lower costs per episode all while maintaining comparable or better patient outcomes.

00:16:14: The return-on investment is proven.

00:16:16: Proven!

00:16:17: So my ultimate question to you is if clinical evidence has closed and economics makes sense what's the bottleneck?

00:16:23: Why isn't this universally adopted standard practice tomorrow?

00:16:27: Ah, well because the barrier is no longer the clinical evidence.

00:16:30: The Barrier Is Systemic Ecosystem Fragmentation.

00:16:33: Explain that!

00:16:34: The technology is entirely ready but the administrative orchestration is lagging years behind.

00:16:40: Sigrid Berge van Rogen highlighted this systemic failure perfectly while discussing the European Digital Health Tech Conference.

00:16:48: She pointed out that incredible innovation struggles to scale, especially in a market like Europe because you aren't plugging into one cohesive healthcare system.

00:16:57: Right it's a patchwork

00:16:58: Exactly!

00:16:59: You are navigating dozens of different reimbursement pathways entirely fragmented legal policies and isolated regulatory bodies.

00:17:07: You can invent the most cost-effective virtual care platform or the most precise robotic system in world.

00:17:13: But if local reimbursement authorities and legacy hospital IT systems aren't aligned to integrate it, simply cannot scale!

00:17:20: It is exactly like having a state of art high speed bullet train... But the tracks across the continent are built at five different widths.

00:17:28: The train works perfectly, but it can't leave this

00:17:30: station.".

00:17:30: That

00:17:31: is reality of health tech right now.

00:17:32: Execution requires navigating the tracks.

00:17:35: we have not the tracks that wish had.

00:17:37: Wow!

00:17:38: So what does all mean for you as a listener?

00:17:40: As you navigate this incredibly complex space...

00:17:43: I think the overarching lessons synthesizing all of these sources is clear.

00:17:48: Whether you are building ambient AI clinical agents, designing public health data federations or rolling out the next generation of modular robotic surgery... The technology alone isn't your product!

00:18:00: No not at all.

00:18:01: Your

00:18:01: product is the workflow.

00:18:02: your product is the governance structure.

00:18:04: Your product is human integration, deeply understanding that Susan Layer of how care is actually physically delivered on the

00:18:12: floor.".

00:18:13: And as we wrap up I want to leave you with a final somewhat provocative thought that builds onto very insightful post by Mathias Goyen.

00:18:19: Oh

00:18:20: yeah this was good one!

00:18:20: Yeah... We have spent an entire deep dive talking about optimizing workflows capturing billions data points at home and measuring operational efficiency but go and warns of a creeping paradox in our industry.

00:18:34: As are digital dashboards become louder than human dialogue, we risk measuring clicks while completely ignoring

00:18:40: purpose.".

00:18:40: That's a scary thought it is.

00:18:42: AI & Robotics can absolutely help clinicians work faster safer with less physical strain.

00:18:49: But if health tech leaders don't intentionally design these systems to protect the joy, psychological safety and fundamental purpose of practicing medicine no amount of workflow optimization will fix clinician burnout.

00:19:01: So as you build a future of HealthTech You really have ask yourself are measuring efficiency at cost empathy?

00:19:08: If you enjoyed this episode, new episodes drop every two weeks.

00:19:12: Also check out our other editions on ICT and Tech Insights, DefenseTech, Cloud, Digital Products & Services, Artificial Intelligence and Sustainability and Green ICT.

00:19:21: Thank You so much for joining us in the deep dive into the real mechanics of HealthTech And don't forget to subscribe So that never miss an update.

00:19:27: We will see ya next time.