Best of LinkedIn: Health Tech CW 06/ 07

00:00:00: This episode is provided by Thomas Allgeier and Franis, based on the most relevant LinkedIn posts on health tech in CW six-and-seven.

00:00:08: Franus equips product and strategy teams with market and competitive intelligence to navigate the HealthTech landscape.

00:00:15: Welcome back to The Deep Dive!

00:00:16: We are looking at the Health Tech Landscape for Weeks Six & Seven of twenty-twenty-six

00:00:22: And I have say going through sources this time...the feeling was Different

00:00:26: different how

00:00:27: well.

00:00:28: usually there's a lot of you know science fiction Cool concepts that are maybe five ten years out right

00:00:34: the flying cars of health care

00:00:35: exactly, but this week It didn't feel like a science fair.

00:00:39: If felt more like a construction site

00:00:42: That's a perfect way to put it.

00:00:43: if there's one big takeaway is this The industry is moving from experimentation to scale the execution.

00:00:49: The sandbox phase is over?

00:00:50: It is!

00:00:51: We're not just seeing cool pilots, we are seeing regulatory-grade platforms actual reimbursement industrial maturity...the plumbing has been laid down

00:00:59: And that's what were here to unpack..we've grouped updates into well four big themes.

00:01:03: Were gonna talk about AI becoming agintic which was a term we definitely need to define.

00:01:08: Then will get in some uh literal superpowers and imaging diagnostics

00:01:13: New wave robotics.

00:01:14: then Big one The money question Digital therapeutics actually getting paid.

00:01:20: It's a packed agenda, so let's get right into it.

00:01:22: Let's

00:01:22: start with AI.

00:01:23: I mean we say AI is changing health care all the time.

00:01:26: That's almost a cliche

00:01:28: but opposed from Effie glow reporting from WA checks in Dubai She used to phrase that really stuck with me

00:01:35: was at.

00:01:35: she said AI is graduating from a chatbot to a co-worker.

00:01:39: And that is the critical shift, she was talking about agentic AI.

00:01:42: right and for anyone listening this is a real leap forward from the generative AI we've all gotten used.

00:01:48: so it's not just chat gpt for doctors.

00:01:50: no

00:01:50: i mean generative ai is passive.

00:01:52: you ask a question it gives you an answer?

00:01:54: It waits!

00:01:55: Agenic Ai has agency its designed to execute tasks

00:02:00: autonomously.

00:02:01: So does it summarize meeting notes

00:02:03: It schedules the follow-up, it updates the patient file.

00:02:06: It flags a prescription conflict all without you holding its

00:02:10: hand.".

00:02:11: And she gave that example from the UAE Ministry of Health.

00:02:14: they have projects where AI agents are handling what was it?

00:02:18: A third of the admin burden...

00:02:19: ...a third and think about what that really means.

00:02:21: this isn't a doctor checking the AI's work every few seconds.

00:02:25: No!

00:02:25: This is AI doing triage AI doing scheduling, its

00:02:29: infrastructure.

00:02:30: and Effie's point about geography was interesting too.

00:02:32: The health care lab for this.

00:02:35: It's not Silicon Valley anymore

00:02:37: No it's the GCC the Gulf region in China.

00:02:39: They're just operationalizing this tech so much faster

00:02:42: which is a huge wake-up call.

00:02:43: I mean if you're only building for Western regulatory speeds You might be building for yesterday

00:02:48: And speaking of the UAE Said Al Hasan shared an update on Amal the

00:02:52: first AI physician assistant.

00:02:54: right And the language they use is key.

00:02:57: It's to augment decision-making,

00:02:59: Augment not replace

00:03:01: exactly.

00:03:01: it reduces the cognitive load But yeah and there's always a but when you move this fast

00:03:06: There are new risks

00:03:07: and This Is where we need To look at what's happening in Italy?

00:03:10: Chiara Galiz had A really sharp analysis on this.

00:03:13: On

00:03:13: The MIA system right correct?

00:03:15: so Italy deployed this generative AI tool mia into fifteen hundred family doctors offices.

00:03:22: Wow That is national scale.

00:03:23: It is and it uses something called RAG, Retrieval Augmented Generation.

00:03:28: Okay break that down for us.

00:03:30: what does that mean here?

00:03:31: So

00:03:31: instead of the AI just you know making things up from the internet, ROGI forces to look at a specific trusted library first

00:03:38: Medical guidelines in this case Right!

00:03:40: It retrieves correct data then generates answer

00:03:44: Which sounds much safer

00:03:46: It is, in theory.

00:03:47: But Chiara points out this very human problem... automation bias!

00:03:51: We trust the machine too much…

00:03:53: We do!

00:03:54: These systems sound so confident—perfect grammar.

00:03:57: A busy doctor asks for a diagnostic frame The AI gives a coherent answer And

00:04:02: the Doctor just accepts it.

00:04:03: and the AI becomes shadow co-author of that medical decision.

00:04:06: That's the scary part.

00:04:08: If the AI misses a nuance Who's responsible?

00:04:11: The doctor signed off but the shadow co author wrote

00:04:14: Confidence is not competence.

00:04:17: And that's why Tim Ruchenbaum's update from Oracle was so relevant,

00:04:20: he was talking about their new AI data platform?

00:04:23: Yes and his point was at this.

00:04:25: agentic intelligence needs deep clean clinical data.

00:04:29: you can't just slap a smart bot on top of a messy system.

00:04:33: it's

00:04:33: an infrastructure play Not Just An App.

00:04:35: You Download

00:04:36: It The Plumbing Of The Hospital.

00:04:38: Okay.

00:04:38: So That'S the Back End.

00:04:39: Let's Shift To The Front End The Eyes Of The Operation imaging and diagnostics.

00:04:44: And looking at the posts from Ryan Fukushima, Dr Daniel Stromer it really does feel like we're getting superpowers!

00:04:50: Let's start with Ryan's insight on page.

00:04:52: predict he used an analogy that just...it clicks

00:04:55: A compressed file.

00:04:56: Yes so think of a standard pathology slide The pink-and-purple tissue sample for cancer diagnosis.

00:05:01: Think if is as a compressed file.

00:05:02: For one hundred years pathologists have looked at this slide seeing cell shapes Morphology.

00:05:07: but the AI sees more.

00:05:09: So much more!

00:05:10: It looks at that same slide and see's molecular data, it decompresses the file.

00:05:14: He said AI can now predict a hundred twenty-three biomarkers across sixteen cancer types

00:05:20: From a single slide.

00:05:21: That

00:05:21: is just wild.

00:05:23: Usually you'd need what?

00:05:23: Expensive genetic sequencing for that

00:05:25: Exactly More tissue, more time A lot more money.

00:05:29: This means You could get that same date Adjust by running code on an image you already have.

00:05:33: It democratizes precision oncology.

00:05:36: But speed is also a superpower, especially in the ER.

00:05:39: And that brings us to The Photon Friday Case.

00:05:42: Cheered by Dr.

00:05:43: Daniel Strowmer This one really hits home.

00:05:45: Chest pain, racing heart, emergency room... The

00:05:48: triple rule out!

00:05:50: This has huge leap into CT tech Usually scanning hearts beating fast as well.

00:05:55: It's a nightmare.

00:05:56: The image gets blurry Right…the

00:05:58: scanner like camera.

00:06:00: If the heart is moving fast, the picture's useless.

00:06:02: But Dr.

00:06:03: Schromer was talking about the semen's naeotom alpha dot prime of photon counting CT Right

00:06:08: and without getting too deep into physics.

00:06:11: these detectors are incredibly fast.

00:06:12: They count individual photons so they can freeze motion

00:06:15: And results.

00:06:16: in this case

00:06:17: One scan one breath hold and diagnosed.

00:06:19: pulmonary embolism aortic dissection coronary artery disease

00:06:23: Three deadly conditions at once.

00:06:25: That difference between innovation and execution.

00:06:28: Execution We saved this person's life in five minutes and the ER.

00:06:32: Incredible, but the tech is only half of battle.

00:06:35: Andrei Hartung added another layer to it

00:06:37: The unified cockpit.

00:06:38: This about human factor.

00:06:40: we're drowning radiologists data.

00:06:42: Right.

00:06:43: if you give them better images with a harder workflow they burn out.

00:06:46: Andre was talking how AI needs reduce scan times put all that data into one clean view

00:06:52: Producing cognitive load

00:06:53: Just as important image resolution.

00:06:56: And Mark Stoffels from Phillips mentioned their acquisition of Spectrowave, which is about combining imaging and AI inside the veins.

00:07:04: So you can decide guide-and-treat all in one.

00:07:06: go closing The Loop?

00:07:08: Exactly!

00:07:09: Okay so speaking of closing A Loop let's talk about robotics when the code meets the physical

00:07:14: world.

00:07:15: We have to start with this story shared by Dr Martha Buchenfeld About a man named Tucker

00:07:20: Moore.

00:07:20: Such a powerful story.

00:07:21: it grounds All This Tech In Reality.

00:07:24: So Tucker's a marathon runner.

00:07:25: He has a freak accident, falls downstairs loses a hand-prit sized piece of his skull

00:07:30: and suddenly His life is on hold wearing a helmet everywhere.

00:07:34: And the standard solution?

00:07:35: A custom titanium implant costs what twenty thousand dollars or

00:07:39: more.

00:07:39: for a lot Of people that's just out of reach.

00:07:41: but dr.

00:07:42: Bokenfeld highlighted The shift to three d printing.

00:07:45: they used a peak implant.

00:07:47: peak Is a high performance polymer.

00:07:49: it's biocompatible.

00:07:51: its light.

00:07:52: This is the crucial part.

00:07:53: It can be three d printed very cheaply.

00:07:55: how cheaply

00:07:56: she said.

00:07:56: in some pilots They're doing it for as low as six hundred dollars

00:07:59: from twenty thousand to six hundred.

00:08:01: that Is democratization and The end of the story.

00:08:05: Tucker ran the nyc marathon just months later.

00:08:08: That's

00:08:08: what scaled execution looks like,

00:08:11: but the big players are moving here too.

00:08:13: we saw updates From Medtronic Raji Kamal on the Hugo system, Jeff Martha on Stealth XIS.

00:08:20: And the trend there seems to be ecosystem?

00:08:22: It is!

00:08:23: Medtronic isn't just selling a robot arm — they're unifying planning navigation and robotics into one workflow...

00:08:29: ...and real-time data's coming in that work flow now too….

00:08:32: Yes

00:08:33: Alice Warden from GE Health Cow posted about integrating their ultrasound directly into robotic surgery

00:08:39: Which is huge because the surgeon is at a console ten feet away.

00:08:42: They lose that sense of touch.

00:08:44: They're flying by vision alone, but adding ultrasound lets you see inside the organ before you cut.

00:08:49: it's like having x-ray vision during operation

00:08:52: and looking ahead.

00:08:53: bird van mirrors from Phillips mentioned this seismic consortium.

00:08:56: over twenty three million euros in EU funding.

00:08:59: This is focused on the brain on minimally invasive treatments.

00:09:03: It's the final frontier for this kind of tech.

00:09:05: Okay so we have smart agents with supervision We have robots But none if matters If no one pays.

00:09:11: which brings us to theme four, digital therapeutics and virtual care.

00:09:15: And we have a major reimbursement milestone.

00:09:19: Lucy Jones shared the news about OvevaDirect in Germany.

00:09:22: This is huge, it's Germany first reimbursed.

00:09:25: digital therapy a digae for both hypertension and weight management

00:09:29: Reimbursed that's the magic word

00:09:31: It is.

00:09:31: there are thousands of wellness apps but oveva showed the data an eleven millimeter hg drop in systolic blood pressure over four percent weight loss.

00:09:40: So because they have hard clinical data The German health system says okay we'll pay for this like a pill.

00:09:45: Exactly, but the regulation around this is tricky.

00:09:48: Asincon analyzed the FDA's twenty-twenty six guidance and it looks like they're opening a side door

00:09:54: for general wellness devices.

00:09:56: right your smartwatches your rings Devices that measure BP's PO to glucose.

00:10:01: So what does the shift?

00:10:02: The FDA saying if you stay in that wellness lane And you don't claim to diagnose or treat a disease You can bypass the strict five-ten K clearance.

00:10:12: So faster market access for manufacturers?

00:10:14: Much faster, but they have to be so careful with marketing if they cross the line into diagnosis there in big trouble and

00:10:20: when you combine that with gamification.

00:10:22: what Mengcha shared about loony health in Singapore is fascinating.

00:10:27: my favorite example of prevention as infrastructure a partnership between Apple and the Singaporean government

00:10:32: And this wasn't a small pilot

00:10:33: no.

00:10:37: and by gamifying, health-closing your rings or any points they saw sustained activity gains.

00:10:42: Gains that could actually reduce mortality risk

00:10:44: By three to thirteen percent.

00:10:45: it proves.

00:10:46: if you design the incentives right at a national scale You can change the health profile of country

00:10:51: And its not just general wellness.

00:10:54: Sina Esimiri reported on dental monitoring securing a hundred million dollars for

00:10:59: remote orthodontic care.

00:11:00: It just proves the model works.

00:11:01: You don't need to go to the dentist every two weeks if an AI can check your selfie video.

00:11:05: Okay,

00:11:05: so let's zoom out.

00:11:06: we have the tech We have the scanners we even have some reimbursement

00:11:10: but you have to play Dell's advocate.

00:11:12: I do.

00:11:13: If this stuff is so good Why isn't it everywhere?

00:11:16: why am i still faxing things to my doctor?

00:11:19: That's the multi-billion dollar question.

00:11:22: And it brings us to our final theme, strategy and the.

00:11:25: so what?

00:11:26: Vladislav Puzhankov had a brilliant take on this.

00:11:29: Demand fulfillment versus demand creation.

00:11:33: Every health tech founder needs to hear this.

00:11:35: he argues that innovation fails if you just try to fulfill demand like answering a hospital's RFP

00:11:40: because The Hospital Just asked for A faster horse slightly better version of What they already have.

00:11:45: precisely Puzonkov used intuitive surgical as an example.

00:11:49: They didn't sell a robot, they created the category of robotic surgery.

00:11:52: They had to build the demand from zero.

00:11:54: If they'd waited for a tender or surgical robot then you would be gone.

00:11:58: You have teach them why it needs your solution.

00:12:01: But even when that demand is there Adoption is still uneven.

00:12:05: That report Jan Beger shared on US hospitals was... A bit of reality check.

00:12:10: It was only forty.

00:12:12: nine percent use predictive AI Less than half.

00:12:16: And it's clustered in wealthy urban areas with cold spots, in rural areas.

00:12:20: So the places that need efficiency most are least likely to have it.

00:12:23: The infrastructure gap is real

00:12:25: Which brings us a final point from Simon Philip Ross at GE Healthcare.

00:12:30: He basically said... No!

00:12:34: The barrier is status quo.

00:12:36: It's our willingness to redesign how we partner and govern these systems.

00:12:40: We had to stop running Pylis

00:12:41: ...and start redesigning this system.

00:12:43: The tech is ready, the question is are we ready to change how we work?

00:12:46: To fit the technology.

00:12:48: That's a challenge isn't it?

00:12:49: moving from the wow phase...to the How Phase and the How Is Messy!

00:12:53: It is but as we saw with the three-D printed skull or the triple rule out scan when We get the HOW RIGHT THE IMPACT IS IT'S IMMEDIATE IT'S PROFOUND

00:13:02: absolutely.

00:13:03: the hard Work of Execution is just beginning.

00:13:05: indeed if you

00:13:06: enjoyed this episode new episodes drop every two weeks.

00:13:10: Also, check out our other editions on ICT and Tech Insights.

00:13:12: DefenseTech Cloud Digital Products & Services Artificial Intelligence And Sustainability in Green ICT.

00:13:20: Thanks for listening!

00:13:21: Don't forget to subscribe so you don't miss the next deep dive.

00:13:24: See ya next time.