Best of LinkedIn: Health Tech CW 38/ 39

00:00:00: This episode is provided by Thomas Allgeier and Frennis.

00:00:03: Based on the most relevant LinkedIn posts on health tech in CW eight and thirty nine, Frennis Eclipse 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: If you are in the health tech space or, you know, maybe just trying to figure out where all that investment money is actually going and scaling up, then this deep dive is definitely for you.

00:00:24: We've sifted through quite a bit of chatter on LinkedIn over the last couple of weeks.

00:00:28: That's calendar weeks, thirty eight and thirty nine to pull out the key strategic insights everyone's talking about.

00:00:33: Yeah.

00:00:34: And the big takeaway of the central theme we kept seeing is this major leap in maturity.

00:00:38: We're really definitively moving past that pilot project proof of concept phase.

00:00:43: The focus now seems to be much more on, okay, how do we aggressively scale this?

00:00:47: How do we embed these technologies right into the clinical workflow?

00:00:51: so they're impacting?

00:00:52: you know, millions of patients like right now.

00:00:54: Right.

00:00:55: So it's less about pure innovation sparkle and more about actual adoption and integration.

00:01:00: We're going to look at AI, obviously, how it's changing things from the ORR table to managing chronic heart conditions.

00:01:07: But there's also this tension we need to talk about.

00:01:09: Oh, yeah.

00:01:10: The faster the tech moves, the harder it seems for the actual health care system and critically patient trust to really keep

00:01:16: up.

00:01:16: Exactly.

00:01:17: So our aim here is to spotlight what's really demonstrably working now.

00:01:22: and maybe even more importantly flag those critical governance and readiness issues.

00:01:28: Because if those aren't tackled, they could become huge roadblocks to adoption down the line.

00:01:32: Okay, let's dive in them, starting with probably the area under the most pressure globally.

00:01:36: Diagnostics.

00:01:38: So our first theme is AI in imaging and radiology, really focusing on those big games and precision and efficiency.

00:01:44: The

00:01:44: pressure is just immense.

00:01:46: We saw Jianji pointing out the situation in England.

00:01:48: Think about this, over one point seven million people waiting for a diagnostic test.

00:01:52: Wow.

00:01:53: That kind of backlog just forces you to look for efficiency everywhere.

00:01:56: And the sources suggest relief isn't just coming from smarter algorithms, but actually smarter hardware too, like next gen CT standards.

00:02:05: Peter Schart from Siemens Health & Years shared some really interesting stuff on Photon Counting CT or PCCT.

00:02:11: Right, PCCT.

00:02:12: It's a fundamental step up in the physics, basically.

00:02:15: It's already giving ultra-high resolution scans and, importantly, significantly lower radiation doses.

00:02:22: But the really strategic piece Peter shared was the scale, this tech.

00:02:26: It's already benefited over two million patients.

00:02:28: Two million?

00:02:29: Okay, that's not a pilot.

00:02:30: No, exactly.

00:02:30: That's a serious, scaled-up rollout with real benefit.

00:02:33: That's

00:02:33: powerful validation.

00:02:35: So that kind of precision... What did they say?

00:02:37: Point two millimeter resolution?

00:02:39: What does that actually mean in practice?

00:02:41: It means incredible clarity and crucially speed.

00:02:45: Andrea Petraca, even noted PCCT, has sort of crossed over into veterinary medicine now.

00:02:50: Huh,

00:02:50: interesting.

00:02:51: Yeah, think about scanning an animal in trauma.

00:02:53: that's seventy-eight centimeter per second speed.

00:02:55: It can sometimes mean they don't need full sedation.

00:02:58: Now, you can save time and risk there.

00:03:00: Imagine the impact in a busy human ER.

00:03:02: That's a really good point about operational efficiency.

00:03:05: But AI is also making strides in the post-processing side, right?

00:03:08: Particularly with MRI.

00:03:09: Absolutely.

00:03:10: Kilian Solte reported on a study using deep learning post-processing on something called black blood MRI, which is already a technique used to spot small, tricky lesions.

00:03:20: What the deep learning model did was boost the sensitivity for finding tiny brain metastases, especially those under five millimeters.

00:03:27: And here's the kicker.

00:03:28: Strategically speaking.

00:03:30: The detection rate jumped from seventy-nine percent up to eighty-six percent.

00:03:33: That's a solid seven percent gain.

00:03:35: Achieved without adding any extra time for the patient in the scanner.

00:03:39: Zero extra scan time.

00:03:41: And that shifts the investment conversation, doesn't it?

00:03:43: It's not just about buying a new expensive machine, the CAPEX.

00:03:46: It's also about the software licensing, the integration and the OPEX side, which, as Roland Roth pointed out, is critical if AI is going to help solve that burnout problem in radiology, where scan demand is just swamping staffing levels.

00:03:58: Yeah, it kind of redefines where value gets created.

00:04:01: And Matthias Goyen from GE Healthcare really painted the bigger picture here.

00:04:04: He sees imaging moving away from being the standalone silo.

00:04:09: Towards being the absolute center of care.

00:04:11: Seamlessly connected to genomics, labs, the EHR, with AI sort of.

00:04:17: predicting issues and prioritizing workflows, that sounds like the future hospital model taking shape.

00:04:22: Which is a perfect segue, actually, from optimizing diagnosis straight into optimizing the intervention.

00:04:28: Let's talk thing two, surgical robotics, or our AI, and even digital twins.

00:04:32: It

00:04:32: really sounds like AI is graduating from being just a guide or making suggestions to becoming, well... almost like another member of the surgical team.

00:04:40: It

00:04:40: really is fundamentally changing roles in the OR.

00:04:42: We saw a pretty vivid example from Dina Ibrahim.

00:04:45: She highlighted an AI-guided autonomous camera system used down in Chile.

00:04:49: Get this.

00:04:50: It allowed a surgeon to perform a single-handed cholecystectomy.

00:04:53: The AI managed the camera entirely.

00:04:55: A single-handed surgery?

00:04:57: That's incredible automation.

00:04:59: But I like that.

00:04:59: Dina immediately added the caution, you know, that the core human stuff, judgment, experience, empathy, that's still irreplaceable.

00:05:06: Absolutely.

00:05:07: Critical balance.

00:05:07: Precision automation meets human oversight.

00:05:10: It's a high wire

00:05:11: act.

00:05:12: And it's not just isolated experiments, is it?

00:05:15: No way.

00:05:16: Mike Marinero reported that Medtronic is actually doubling its presence in London.

00:05:20: They're creating their biggest global hub there.

00:05:23: specifically for AI, digital tech, and surgical robotics.

00:05:27: London specifically.

00:05:29: Is that about talent?

00:05:31: Or maybe regulation?

00:05:32: Probably a mix, but talent and infrastructure are likely key drivers.

00:05:36: It signals a huge investment in centralizing R&D and strategy for these really complex platforms.

00:05:41: And Julio Zappello added that the conversations happening at Medtronic events like their Hugo Lennon S. Forum are all about accessibility now, how to get robotic surgery beyond the big academic centers and into more community hospitals.

00:05:52: Making it more mainstream.

00:05:53: And beyond the actual robots, we're seeing progress in planning tools too, like digital twins.

00:05:59: Yeah, Oliver Fring shared work on a liver digital twin.

00:06:02: The idea is to predict the outcomes of thermal ablation before you even start the procedure.

00:06:07: That work got recognized at MICCI.

00:06:10: So a digital twin is like a personalized simulator for that specific patient's liver.

00:06:14: Exactly.

00:06:15: Being able to model and predict how that complex ablation will go for that individual, it massively reduces risk and improves the planning.

00:06:22: It shows real practical steps towards safer, truly personalized surgery.

00:06:27: Okay, so moving from those sort of episodic surgical interventions, let's shift to chronic care.

00:06:32: Specifically high impact areas like cardiology, theme three, cardiology and chronic care at scale.

00:06:38: This seems to be where big partnerships are really accelerating things.

00:06:41: Definitely.

00:06:42: The biggest news here was around Phillips and Optum Health Care.

00:06:44: Alison Jones, Diego Carleso, Julia Strandberg, several people highlighted this huge national partnership in the U.S.

00:06:51: It involves integrating Phillips with cardiac monitoring tech, their mobile cardiac telemetry, or MCOT, and the E-PASH devices directly into Optum's network.

00:06:59: That's massive reach.

00:07:01: What was it?

00:07:01: Three point four million members across twenty two states.

00:07:04: So what's the operational advantage there?

00:07:06: Just easier access.

00:07:07: It's more than access.

00:07:08: It really streamlines the whole referral pathway.

00:07:12: Julia Stranberg made the point that it gives clinicians actionable insights almost in real time from that remote monitoring.

00:07:18: So instead of data getting stuck in silos, cardiologists, primary care docs, they get this continuous feed.

00:07:24: It enables truly proactive care models at a really significant scale.

00:07:28: But putting AI into cardiac care must come with uniquely high stakes, right?

00:07:33: Trust has got to be paramount.

00:07:34: Absolutely.

00:07:35: Anurata Bambawal shared some key insights from ESC twenty twenty five, emphasizing that for AI to really take hold in cardiology, it needs what she called credible AI.

00:07:45: Credible AI.

00:07:46: OK, what does that mean?

00:07:47: Well, three core things.

00:07:48: First, it has to solve a real clinical problem, not just be tech for tech sick.

00:07:52: Second, it needs to be built on properly validated, high-quality data.

00:07:55: Garbage in, garbage out is amplified here.

00:07:57: And third, and this is crucial, it must reinforce the physician-patient relationship, not undermine it.

00:08:03: In heart care, that trust is just non-negotiable.

00:08:06: Makes sense.

00:08:08: And speaking of solving real problems, Jan Beger talked about

00:08:11: A.I.'s

00:08:11: potential in diagnosing hypertrophic cardiomyopathy, HCM.

00:08:16: Yeah,

00:08:17: HCM can be quite tricky to diagnose accurately, but the A.I.

00:08:20: models Jan discussed showed they could differentiate it from other conditions with really high accuracy.

00:08:26: We're talking area under the curve, AUC, up to point nine eight.

00:08:30: Almost perfect then.

00:08:31: In that specific study group, yes.

00:08:34: The big challenge, as always, is ensuring that accuracy holds up when you generalize it across diverse real-world populations.

00:08:42: That's the next step.

00:08:43: And sticking with chronic conditions, digital approaches are proving effective in weight management too.

00:08:47: They are.

00:08:47: Martin Fidok reported on services like OVIVA's digital weight management program.

00:08:51: When combined with medication, patients achieved an average weight loss of eight point five percent of their starting weight at six months.

00:08:57: Eight

00:08:57: point five percent is clinically significant.

00:09:00: It absolutely is.

00:09:01: It validates that these digital first solutions can work for conditions needing long term behavioral change.

00:09:06: It shows if you can scale the platform effectively, you can scale the clinical outcomes too.

00:09:10: OK, so we've covered a lot of ground on the tech momentum, the scaled deployments, incredible progress.

00:09:16: But now we have to hit those friction points.

00:09:18: The system challenges that, if they're not sorted out, could really slow things down.

00:09:24: This brings us to our final theme, ethical gaps and system readiness.

00:09:28: Yeah, this is where that tension we mentioned earlier really comes into sharp focus.

00:09:31: When innovation moves this fast, the ethical frameworks, the governance, and crucially, patient trust, they often struggle to keep pace.

00:09:40: Segerd-Berge van Roysen made some great points about why addressing these gaps is so vital for safety and equity.

00:09:46: And the sources paint a picture of a pretty significant disconnect, don't they?

00:09:50: Joe Bell-Fontaine and Juan T. referenced that Phillips Future Health Index report from Canada.

00:09:55: Right.

00:09:55: Get this.

00:09:56: Eighty-six percent of healthcare professionals are optimistic about AI.

00:10:01: But only forty-nine percent of patients feel the same way.

00:10:04: That gap, thirty-seven points, that's a massive hurdle for adoption.

00:10:07: Think about it.

00:10:08: For every ten clinicians eager to use a new AI tool, nearly four of their patients are skeptical or actively distrustful.

00:10:15: You just can't scale effectively with that kind of mismatch.

00:10:18: And the patient worries are specific, aren't they?

00:10:21: Sigrid noted, fifty-two percent worry about AI in privacy.

00:10:25: And even more, sixty-four percent see potential bias based on ethnicity as a major health issue with AI.

00:10:31: Yeah, and importantly, Sigrid didn't just flag the problems, she offered concrete strategies.

00:10:36: Things like... actively adjusting patient selection criteria for trials, making sure data collection is truly representative, even building fairness constraints right into the algorithms themselves.

00:10:46: Actionable steps.

00:10:47: And on the clinician side, that old debate about AID skilling popped up again.

00:10:51: It always does.

00:10:53: But Austin Chang had a really good take on it.

00:10:54: He basically reframed it not as de-skilling, but as evolution.

00:10:58: His point was that medical innovation always changes skill sets.

00:11:02: Think about cross-sectional imaging, replacing older techniques or laparoscopy.

00:11:06: changing surgery.

00:11:07: Old skills get replaced by newer ones that enable better, safer, more efficient care.

00:11:11: It's evolution.

00:11:12: That's

00:11:12: a great way to put it.

00:11:13: But the challenge now isn't just adapting skills, it's also managing the sheer volume of data, right?

00:11:18: Vicki Britton highlighted this issue of patients feeling overwhelmed.

00:11:21: Totally.

00:11:22: Patients are drowning in data wearables, chatbots, Dr.

00:11:25: Google, social media, but often without any real framework to make sense of it

00:11:30: all.

00:11:30: So the answer isn't necessarily less data, but better tools to interpret it.

00:11:34: Exactly.

00:11:35: And Yossi Matias presented a possible solution, a wayfinding AI agent.

00:11:40: based on Google's Gemini.

00:11:41: The key innovation was using proactive guidance and being transparent about its reasoning.

00:11:46: Showing its work, basically.

00:11:48: Pretty much.

00:11:48: And that transparency seemed to build trust.

00:11:51: It led to conversations that were significantly longer and rated as more helpful by users almost five turns on average compared to just over three for a standard AI.

00:12:00: Patients engaged more when they understood the why.

00:12:03: Which brings us to the final kind of paradoxical point about the system itself.

00:12:07: Bill Russell summed it up powerfully.

00:12:09: Despite all this enthusiasm and tech push, only thirteen percent of healthcare organizations actually feel ready for the demands AI will place on

00:12:16: them.

00:12:17: Thirteen percent.

00:12:18: Yet, paradoxically, Russell also found that sixty-two percent of these same organizations are prioritizing hiring people with AI skills over those with traditional healthcare experience.

00:12:28: Wow.

00:12:29: So the desire is there, the hiring focus is shifting?

00:12:32: The actual organizational readiness, the infrastructure, the ethics frameworks, the workflows, it's seriously lagging behind the ambition.

00:12:39: It's a major disconnect.

00:12:40: So pulling it all together, we've seen incredible acceleration in real world scaling and precision tech, PCC, deep learning and MRI, surgical robots, massive cardiac care partnerships, huge momentum.

00:12:53: But the bottleneck, the strategic gatekeeper really seems to be the human element, bridging that patient trust gap and tackling the system's readiness deficit.

00:13:02: That tension is absolutely the core takeaway from the sources these past two weeks.

00:13:07: You've got AI getting deeply embedded in diagnostics, even assisting in surgery.

00:13:11: The speed is undeniable.

00:13:12: So the key question for every organization in this space has to be, how do you balance that urgent need for rapid scale deployment with the absolute non-negotiable requirement for strong ethical governance and, crucially, earning and maintaining genuine patient trust?

00:13:27: That's the tightrope everyone's walking right now.

00:13:29: The perfect question to leave our listeners with, if you enjoyed this deep dive, new episodes drop every two weeks.

00:13:35: Also check out our other editions on ICT and tech insights, defense tech, cloud, digital products and services, artificial intelligence and sustainability and green ICT.

00:13:44: Thank you for joining us for this focused look at the health tech ecosystem.

00:13:48: Don't forget to subscribe and we'll catch you on the next deep dive.