Best of LinkedIn: Health Tech CW 36/ 37

00:00:00: This deep dive is provided by Thomas Allgayer and Frennus, based on the most relevant LinkedIn posts on health tech in CW, three, six, three, seven.

00:00:09: Frennus equips product and strategy teams with market and competitive intelligence to navigate the health tech landscape.

00:00:16: And yeah, over the past couple of weeks, LinkedIn has just been absolutely buzzing with some really significant trends in health tech.

00:00:21: We've sifted through it all and pulled out the key insights for you.

00:00:24: That's right.

00:00:25: We're aiming to, you know, cut through the noise here, give you a shortcut to being genuinely well informed on the top health tech trends and insights from calendar weeks, thirty six and thirty seven.

00:00:35: We'll be looking at, well, everything from AI integration and governance challenges to the big market shifts, really helping you get a handle on not just what's happening, but you know, what it actually means for the industry.

00:00:45: Okay, great.

00:00:46: Let's dive right in then.

00:00:48: Our first big theme seems to be AI strategy, regulation and trust.

00:00:52: What really stood out is this clear emphasis from leaders.

00:00:56: seeing AI as a clinical amplifier, you know, something to solve real problems.

00:01:01: But, and this is crucial, it's paired with a very strong focus on governance, accountability, and well, building trust.

00:01:09: It feels like the conversations definitely move past just capabilities.

00:01:12: Yeah, absolutely.

00:01:14: And it's how people are framing it.

00:01:15: Matthias Guayan, reflecting on WHX Tech Dubai, put it really well.

00:01:19: He stressed that AI adoption and health care, it's not about these isolated pilots anymore.

00:01:23: It needs deep integration into existing workflows.

00:01:26: And he made this compelling case for really rigorous regulation, essentially treating AI like we treat medicine.

00:01:32: Building that trust, you know, through transparency with patients and clinicians, that's paramount for him.

00:01:37: And AI is augmentation, right?

00:01:38: Not substitution, the final call.

00:01:39: always with a physician, like a co-pilot.

00:01:41: That

00:01:41: co-pilot analogy makes a lot of sense, especially when you look at the regulatory debates heating up, particularly in oncology and diagnostics.

00:01:49: There's a real push to update medical device rules, strengthen testing, make sure these things are reliable.

00:01:56: Elena Abitoff's post really hammered home that interpretability is just fundamental for transparency, accountability, fairness, all of it.

00:02:05: She pointed out, you know, the FDA has approved over seven hundred AI tools, but adoption.

00:02:09: It stalls if there isn't that basic trust and understanding.

00:02:12: Approvals in enough you need buy-in.

00:02:13: Exactly.

00:02:14: And that leads right into maybe an even trickier bit.

00:02:17: Accountability.

00:02:18: Jan Beger raised this really important point about the lines blurring.

00:02:21: I mean, if AI shapes a decision, who is responsible?

00:02:24: And it gets even murkier when these models update themselves silently in the background.

00:02:28: Transparency, like he said, it's just non-negotiable when the tool itself can change without you even knowing.

00:02:34: It's a big legal and ethical knot we're still entangling.

00:02:37: It really feels like we're moving past the initial AI hype cycle, doesn't it?

00:02:41: The commentary now stresses, you know, less speculation, more pragmatic value.

00:02:46: Things like inclusive AI built on diverse data.

00:02:49: Brooke Kosic made a great point about just focusing on solving actual clinical problems, prioritizing patient needs, building tools that actually remove tasks from clinicians, not just pile more on.

00:03:01: That's where the real wins are.

00:03:02: Right.

00:03:03: And, uh, speaking of practical ethics, Sigurd Burgband-Roygen laid out five big mistakes to avoid with LLMs, large language models, things like ignoring ethics, overlooking bias, skipping human oversight, disregarding privacy, underestimating misinformation risks.

00:03:18: It's basically a checklist for deploying these things thoughtfully.

00:03:21: Yeah, a really useful framework.

00:03:22: And on safeguards, Sarah Gebauer shared insights from a chat with Morgan Jeffries, who's a medical director of AI.

00:03:27: Morgan really emphasized that FDA approval isn't some kind of green light to just set it and forget it.

00:03:32: You need constant post-deployment monitoring.

00:03:35: And AI literacy for everyone involved, clinicians and patients, that's becoming the ultimate safety net.

00:03:40: It's about ongoing vigilance.

00:03:41: Such a vital point.

00:03:43: Because real Fagnar flagged a serious worry, doctors potentially using unapproved AI, that poses real safety risks.

00:03:51: He reminded everyone that solid regulations from bodies like the IMDRF, UMDR, FDA, Health Canada, Australia's TGA, they're already there.

00:04:00: They need to be followed.

00:04:01: Innovation is great, but it has to be anchored in safety and compliance.

00:04:05: Simple as that.

00:04:05: Okay,

00:04:06: so if the big AI theme is this shift towards responsible, integrated, trustworthy applications.

00:04:11: Let's switch gears.

00:04:12: Let's look at clinical care, diagnostics, and monitoring.

00:04:14: We're seeing some really fascinating progress here.

00:04:17: Cardiology seems to be a major test bet.

00:04:19: Yeah, cardiology is definitely lighting up.

00:04:20: What's really striking, though, is this Philips Future Health Index snapshot, Carla Goulart Perron, and Mark Stauffles mentioned it.

00:04:27: It shows this interesting gap, clinicians.

00:04:30: super optimistic about AI and cardiology, like, eighty-four, ninety-three percent patients.

00:04:35: Much more hesitant, only fifty-six percent optimistic.

00:04:38: That's a clear trust gap the industry needs to address.

00:04:40: And Claire Sattler, DeSoussa, Ibrito added that something like seventy-nine percent of cardiac care pros lose valuable time because of incomplete patient data.

00:04:49: So you can see why AI solutions are desperately needed there.

00:04:52: That inefficiency is a huge target for AI.

00:04:54: And Ryan Fukushima shared news about Tepispixel getting FDA-FiveTenK clearance for enhanced cardiac imaging.

00:05:00: That clearance means it's substantially equivalent to existing tech, right?

00:05:03: Their goal seems clear.

00:05:04: Provide more complete info, integrate different data types, basically move medicine from subjective art towards, well, more precise science.

00:05:11: Getting the full picture.

00:05:12: And diagnostic AI generally is really daining steam.

00:05:16: Komeil Nasrelahi mentioned Tempest AI using Page's massive pathology slide library acquired recently to seriously improve cancer diagnosis.

00:05:25: That's leveraging huge data sets.

00:05:27: And outside cancer, cited health just secured, was it, forty four million dollars for a platform to detect esophageal disease early.

00:05:35: So real progress in key diagnostic areas.

00:05:38: And on the pure innovation front, this is pretty cool.

00:05:40: Robert Slippin talked about continuous protein and inflammation monitoring.

00:05:44: Think CGM, but for way more biomarkers.

00:05:47: He even called it potentially the most important advance yet for precision health.

00:05:51: It's powered by these wearable biosensors from Shayna Kelly's team.

00:05:54: I mean, imagine the proactive health possibilities.

00:05:57: Absolutely transformative potential there.

00:05:59: And then Pira Fattacari showed that the FDA cleared the Apple Watch for hypertension detection.

00:06:03: Just think about the impact of an everyday wearable, flagging, high blood pressure.

00:06:07: It could be huge for cardiovascular and renal health.

00:06:10: enabling earlier intervention, bringing powerful insights right to the individual seamlessly.

00:06:16: Which all comes back to getting clinicians on board, right?

00:06:19: And that seems tied to tools that actually reduce their workload, cut down on alarm fatigue, and just fit into their day.

00:06:26: Patrick Mann's highlighted how Philip CMU, for example, cuts through the noise, prioritizes alerts, helps create a calmer care environment.

00:06:33: It tackles that crazy statistic up to three hundred and fifty alarms per patient per day.

00:06:38: Imagine relieving that cognitive load.

00:06:40: Yeah, that's huge.

00:06:41: Federa Company has echoed that the really impactful AI apps right now are things like medical imaging, diagnostics, automating clinical documentation, which is big burnout fighter, and risk prediction alerts.

00:06:53: The more complex stuff is still much but these foundational uses, they're making a difference now.

00:06:59: So what's the next step for AI in the clinic?

00:07:01: Youngcha's point was interesting.

00:07:03: He said, current AI handles sort of slices of cognition, specific tasks, but a future AI doctor needs to integrate everything orchestrated across the whole job from notes to executing rules to much more holistic vision.

00:07:15: And for tricky diagnoses, Jacob Plummer noted that explainable AI.

00:07:19: like Aida Health, can actually slash the time it takes to get a correct diagnosis for complex conditions globally.

00:07:26: Connecting patients to the right therapy faster, even in top health systems, that explainable part is key for trust.

00:07:33: Definitely.

00:07:34: And then Yolbak has shared that GBT-V hit ninety-three percent accuracy on radiation oncology exams.

00:07:40: Yeah.

00:07:40: That's, well, that's staggering.

00:07:42: significantly better than older versions.

00:07:45: A new benchmark, really.

00:07:46: Though still emphasizing human interaction, of course.

00:07:48: Of course.

00:07:49: Always the human element.

00:07:50: And in surgery, Dr.

00:07:51: Quinlan D. Buchlak reviewed a study on a new way to improve surgical video segmentation, basically helping the AI better identify things during an operation.

00:08:00: This improves guidance, training, safety, especially when the view isn't perfectly clear.

00:08:05: But it's not all straightforward progress.

00:08:06: Daniel Yang brought up a Lancet study that flagged potential de-skilling in GI docs using AI-assisted colonoscopy tools.

00:08:14: That's concerning.

00:08:14: It highlights these potential unintended consequences.

00:08:17: we need to watch really carefully as AI gets embedded deeper.

00:08:20: How do we balance assistance with maintaining core skills?

00:08:23: It's a critical question, and this whole complex, rapidly evolving picture has definitely spurred a lot of corporate action, which brings us to our next theme, corporate moves, investments, and partnerships.

00:08:35: Lots happening here.

00:08:36: Yeah, quite a bit.

00:08:36: Osama Ede noted Oracle putting a hefty one hundred and eighteen million into AI-led vaccine research at Oxford.

00:08:43: That's significant.

00:08:44: It really highlights AI's potential in massive impact areas like antimicrobial resistance.

00:08:50: And it suggests an infrastructure-first approach, building the foundation to scale reliably.

00:08:54: Right.

00:08:54: And Kamal Nasrili, he shared Phillips, is planning over one hundred and fifty million dollars in new investment for MedTech manufacturing and R&D, specifically in the

00:09:02: U.S.,

00:09:02: clearly focused on expanding production of their AI-powered innovations, a big commitment.

00:09:07: We also saw Teladoc acquiring telecare over in Australia.

00:09:10: Jordy Bolcell's canals pointed this out.

00:09:12: It's about strengthening virtual care, right?

00:09:14: Broader specialty access, wider reach, combining strengths.

00:09:18: Matt Rosenberg really framed virtual care as the backbone for whole person integrated care.

00:09:23: makes sense.

00:09:24: And Bill Russell had an interesting take on Epic's dominance.

00:09:27: He argued it's more about strategy than just tech.

00:09:30: I mean, with access to three hundred million patient records, Epic is essentially infrastructure at this point.

00:09:36: That's a powerful position.

00:09:37: Definitely another big one.

00:09:39: Dot Maddox, the scientific informatics firm, being acquired by Siemens for five point one billion dollars.

00:09:45: Comeo Nasroli mentioned this too.

00:09:47: The goal seems to be accelerating AI driven R&D in life sciences, streamlining discovery.

00:09:53: And zooming out globally, Vincenzo Ventrocelli wrote about transformative healthcare innovations across the Middle East, Turkey, and Africa, the common thread, building resilient systems using AI and telehealth.

00:10:05: It's a global push.

00:10:06: That ecosystem approach is key.

00:10:08: Lama Ibrahim and Oracle Cloud highlighted these new alliances popping up, cloud providers, AI vendors, health systems.

00:10:15: all partnering up.

00:10:16: That's how you actually operationalize things like conversational AI and voice workflows in daily practice.

00:10:21: And on the public health side, Steven G noted, verily getting recognized for their work in wastewater epidemiology, using it as a scalable early warning system for infectious diseases.

00:10:31: Super important for preparedness.

00:10:32: Okay, so that brings us nicely to our final theme, workforce design and adoption.

00:10:38: Because ultimately, none of this tech matters if clinicians don't trust it or can't use it effectively, right?

00:10:44: Thoughtful design and implementation seem absolutely critical.

00:10:47: Yeah, and there are some great examples.

00:10:49: Brian Pittman shared how Medtronic changed their onboarding.

00:10:52: They focused reps on just one product category at a time.

00:10:56: Result.

00:10:57: Ramp time dropped from nine months to five.

00:10:59: It just shows, you know, cognitive load is real.

00:11:01: More information doesn't always mean more impact.

00:11:04: Simplicity wins.

00:11:05: That's a brilliant lesson.

00:11:06: And Kieran Daley made such a crucial point.

00:11:09: Build health tech around the patient, not just the clinic.

00:11:12: Because most care happens at home.

00:11:14: In daily life, our digital tools need to reflect and support that reality.

00:11:18: Faye DeWon put it beautifully.

00:11:20: Medical devices are where engineering meets empathy.

00:11:23: The goal.

00:11:24: Restore trust, dignity, and time.

00:11:26: for everyone involved

00:11:27: and it's not just about the shiny new tools.

00:11:30: Spirit and Geek-Eekus Panousis talked about healthcare technology management, shifting from just fixing broken stuff to being proactive using analytics.

00:11:39: That frees up teams to focus more on patients, less on logistics, smarter management of the tech we have.

00:11:45: Right, and Ross Hadfield pointed out that the tech stack in healthcare, it's not invisible anymore.

00:11:49: AI changes how every layer impacts safety, cost, outcomes, so you need resilient AI ready services and clinician friendly design.

00:11:59: Both are essential.

00:12:00: Absolutely.

00:12:01: That usability piece is huge.

00:12:03: Vicki Britton mentioned this digitally literate patient paradox.

00:12:07: Patients are so frustrated with disconnected systems, they're manually uploading lab results to AI chatbots just to try and make sense of it all themselves.

00:12:14: It just screams for simple, integrated, trustworthy solutions.

00:12:18: A major pain point.

00:12:19: And Adam Pellegrini offered a dose of realism.

00:12:22: He reflected that a lot of digital health innovation kind of reinvents the wheel.

00:12:26: He suggested new players should maybe focus on solving fundamental problems from ten years ago that still aren't fixed, instead of just overselling future concepts, deliver real utility now.

00:12:35: That's solid advice.

00:12:36: Simon Philip Roest, featuring GE Healthcare's Scott Miller, emphasized cloud and AI as basically survival tools for radiology, given the workforce shortages and burnout, enabling remote work, standardizing protocols, bringing care closer to patients, it's becoming essential infrastructure.

00:12:54: And looking at Europe, Mark Slade has reported that digital health innovators are pushing the EU for a unified evaluation framework for MedTech.

00:13:01: To fight fragmentation, prevent de-skilling, ensure some technological sovereignty, makes sense.

00:13:06: Yeah, consistency is key for scaling.

00:13:09: Valerie Albert's study on digital health apps showed their potential cost-effectiveness, but she stressed, we need standardized economic evaluations, better trial reporting.

00:13:17: Basically, we need solid evidence to back up the value claims for reimbursement and wider adoption.

00:13:22: And finally, Andrei Krestresovic shared this really powerful vision for AI and mental health, making care TenX better, TenX more accessible.

00:13:31: He sees AI companions as force multipliers, helping patients and providers, but always under professional oversight.

00:13:38: not replacements, huge potential impact there.

00:13:41: So, yeah.

00:13:42: As you can clearly tell, the health tech landscape is just incredibly dynamic.

00:13:45: It's moving fast, but there's also this really crucial ongoing conversation about trust, regulation, and making sure this stuff actually works in practice.

00:13:54: We're moving from pure experimentation towards more structured, responsible scaling.

00:13:58: If you enjoy this deep dive, new deep dives drop every two weeks.

00:14:01: 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.

00:14:10: Thanks for joining us.

00:14:10: Be sure to subscribe so you don't miss our next deep dive.

00:14:13: And maybe

00:14:14: a final thought to leave you with, as Daniel Yang shared, that Lancet Study raised questions about potential de-skilling with AI-assisted colonoscopy.

00:14:22: It makes you wonder, doesn't it?

00:14:24: As we race ahead with AI adoption, how do we consciously design and implement these tools to ensure we are always enhancing critical human skills and clinical judgment rather than inadvertently letting them atrophy?

00:14:35: Something keep thinking about.