Best of LinkedIn: Health Tech CW 46/ 47

00:00:00: This deep dive is provided by Thomas Allgaier and Frennis, based on the most relevant LinkedIn posts on health tech in CW-Forty-Six and Forty-Seven.

00:00:08: Frennis equips product and strategy teams with market and competitive intelligence to navigate the health tech landscape.

00:00:15: Welcome back!

00:00:16: Okay, let's just jump right in.

00:00:18: We spent the last couple of weeks tracking the big health tech trends on LinkedIn, specifically weeks, forty six and forty seven.

00:00:25: Right.

00:00:26: And what was really striking was this this palpable shift in the conversation.

00:00:30: It feels like we're moving past the future hype and we're getting into the nuts and bolts of real embedded clinical tools that are actually changing how things are done today.

00:00:40: That's exactly it.

00:00:41: The mission here is to cut through all that noise.

00:00:44: And we're seeing AI move decisively out of these limited pilots.

00:00:48: into products you can actually deploy.

00:00:50: But that forces you to look at the foundations, things like cloud, data standards, security.

00:00:55: Because without that bedrock, the AI just can't scale.

00:00:58: And it's not just the tech, right?

00:00:59: It's the human element.

00:01:00: We saw so much about how innovation is widening access to care, which is

00:01:04: great.

00:01:06: But at the same time, it's introducing some really serious new questions about clinical rigor and patient safety.

00:01:14: It's like a complex picture.

00:01:16: It is.

00:01:16: So let's start where the change is happening fastest.

00:01:19: The core of diagnosis.

00:01:20: OK.

00:01:21: AI is just fundamentally reshaping radiology.

00:01:24: And the big step now is what they're calling multimodal integration.

00:01:27: So not just looking at one image in isolation.

00:01:31: Exactly.

00:01:32: As Matthias Goyen noted, the future of precision medicine isn't going to come from a scan alone.

00:01:36: You have to combine that imaging with pathology, with genomics, with the whole clinical context.

00:01:42: And that

00:01:43: fusion is getting incredibly specific.

00:01:44: I saw a perfect example of this.

00:01:46: Yoel Baca is highlighted to study on prostate cancer.

00:01:49: They show that when you combine PSMA-PET imaging with AI analysis, it just drastically enhances how you characterize lesions.

00:01:57: Right, and predict patient outcomes.

00:01:59: It's a huge leap over just using the imaging by itself.

00:02:02: So you're getting sharper, earlier insights that a clinician can actually act on.

00:02:05: But to do To handle all that data in real time, you need a massive performance leap in the hardware.

00:02:11: And we saw an answer for that.

00:02:13: TheodoraSaz reported on a breakthrough from Philips and NVIDIA.

00:02:17: With the GPUs?

00:02:18: Yes, using GPU tech to accelerate ultrasound beamforming up to ten times faster with zero loss in image quality.

00:02:25: Ten

00:02:26: times?

00:02:26: I mean, that's a game changer in a busy hospital.

00:02:29: That's not a lab improvement.

00:02:30: That's an operational benefit today.

00:02:33: And the speed has these other impacts I found fascinating.

00:02:35: And hey, Helwitz pointed out that AI acceleration in MRI, like advanced denoising.

00:02:40: Which shortens the scan time.

00:02:41: Exactly.

00:02:42: That directly minimizes the energy footprint of these really high energy systems.

00:02:46: So you're connecting clinical efficiency straight to sustainability.

00:02:49: That's

00:02:50: a perfect example of optimization.

00:02:52: And we also saw this maturing in really specific areas like neuroimaging.

00:02:55: Oh,

00:02:55: yeah.

00:02:56: Luis Cuevas announced Phillips' extended partnership with Cortex.ai.

00:03:01: They're integrating.

00:03:02: AI neuroanalytics directly into the MR systems themselves.

00:03:06: Which means you're moving from a doctor's subjective interpretation of a brain scan to objective quantitative data much faster.

00:03:14: Exactly.

00:03:15: Critical for tracking conditions like dementia over time.

00:03:18: So moving from diagnosis into actual treatment, the conversation got really focused on a high precision intervention.

00:03:25: And the big headliner was digital twins.

00:03:28: And this isn't science fiction anymore.

00:03:30: No, it's being positioned as a really practical bridge from R&D to treating individual patients.

00:03:36: Alina Schvitz had a great post clarifying this.

00:03:38: A digital twin isn't just a scan.

00:03:41: It's a full virtual replica of a patient.

00:03:43: Built

00:03:44: from what?

00:03:44: EHRs?

00:03:45: Genetics?

00:03:46: Everything.

00:03:47: EHR, scans, genetics, wearable data, you put it all together to create these incredibly predictive models.

00:03:53: And they're being used now to simulate drug trials for things like Alzheimer's, which cuts down the sample size needed.

00:03:59: Which is massive.

00:04:00: Think of the time, the cost, the reduction in patient risk, and in surgery, the value is even more immediate.

00:04:05: Right.

00:04:06: Matthew Juliano from Medtronic presented their pipeline for robotic surgery, digital twins.

00:04:11: You use it for capture, playback, training.

00:04:13: So it's basically a flight simulator for a surgeon.

00:04:16: That's a perfect way to put it.

00:04:18: It pairs the robot's precision with simulation to improve safety before you even make an incision.

00:04:23: Newen P called it a force multiplier.

00:04:26: A force multiplier.

00:04:27: I like that.

00:04:28: Yeah, you're using AI and augmented reality to give the surgeon this real-time, three-D anatomical overlay.

00:04:34: Like giving them X revision.

00:04:35: Basically.

00:04:37: And we have concrete examples of this in the field now.

00:04:40: Mark Stauffles and Ruben Olivier highlighted a collaboration between Phillips and Edwards Life Sciences.

00:04:45: On

00:04:45: that hashtag device guide solution.

00:04:47: Yes.

00:04:48: It's designed to simplify a really complex, minimally invasive heart procedure mitral valve repair.

00:04:55: It turns confusing two-D imaging into intuitive three-D guidance for the surgical team.

00:05:00: Which is a huge safety boost for something that requires incredible precision.

00:05:04: But all this amazing tech is only good if people actually use it.

00:05:08: And that means supporting the human workflow, not disrupting it.

00:05:14: He said healthcare needs tech that disappears so compassion can take the front seat.

00:05:17: Right.

00:05:18: If the technology is louder than the patient, you've failed.

00:05:20: So

00:05:20: let's unpack that.

00:05:21: I mean, think about burnout.

00:05:22: The most practical use case we saw was ambient AI scribes.

00:05:26: Krista Calpe showed a study showing that using these AI scribes dropped physician burnout from over fifty percent down to about thirty eight percent.

00:05:33: That's a

00:05:34: huge drop.

00:05:35: The ROI there isn't just saving time on notes, it's keeping your doctors from quitting.

00:05:40: And that gets to the core of adoption.

00:05:42: Sathvik Billakanti pointed this out with that famous

00:05:46: story.

00:05:46: Multi-million dollar AI system versus the WhatsApp

00:05:49: group.

00:05:49: Yeah.

00:05:50: Why did the WhatsApp group win?

00:05:53: Because the tech team built a huge system, but the nurses just needed to solve one daily frustration.

00:05:58: It's

00:05:58: a perfect lesson.

00:05:59: The tech that wins just makes the human better at their job.

00:06:03: And big tech is clearly paying attention.

00:06:05: They

00:06:05: are.

00:06:06: Oracle Health is embedding AI deep into the EHR for note automation.

00:06:10: And Carl Nislow noted that OpenAI is exploring a personal health assistant.

00:06:14: I mean, they have eight hundred million weekly users already asking chat GPT health questions.

00:06:19: Wow.

00:06:19: Shifting gears a bit to access.

00:06:21: We saw some powerful examples of tech closing huge equity gaps.

00:06:25: Jantomar highlighted India's labs on wheels, mobile health vans using IOMT and FiveG to bring diagnostics right into remote villages.

00:06:34: And we're seeing that connectivity finally reached the most remote areas here, too.

00:06:38: Joel Ibarthelamy gave examples of how digital health is transforming indigenous care in North America.

00:06:44: How?

00:06:44: Low Earth orbit satellites.

00:06:46: Well, ELUS are providing the broadband that traditional infrastructure just couldn't deliver.

00:06:51: But that access creates new problems.

00:06:54: We saw a really sharp debate around the new GOP-I weight loss drugs.

00:06:59: Right.

00:07:00: Michelle Davy noted how telemedicine is expanding access, which seems good.

00:07:04: But then Nisha Chellam offered a really strong warning.

00:07:07: She called this democratization a double-edged sword.

00:07:10: Because

00:07:11: you're dispensing powerful drugs transactionally without the comprehensive care and monitoring that you go with them.

00:07:16: Exactly.

00:07:17: You risk serious complications and the drugs don't even work as well over time.

00:07:21: The innovation and access has to be matched with clinical responsibility.

00:07:25: Which

00:07:25: brings us right to the biggest challenge of all.

00:07:28: Why do so many of these heavily funded health tech startups just...

00:07:34: Well, Reza Hosseini-Gomi, who's been building these companies for fifteen years, he laid out the pattern.

00:07:40: They fail because they skip the hard parts.

00:07:42: They don't secure trust, they don't integrate properly, and they can't figure out reimbursement or date equality.

00:07:48: So he's saying healthcare is a trust problem wrapped in regulation.

00:07:51: It's not just a tech problem.

00:07:52: Exactly.

00:07:53: You look at the big failures.

00:07:55: Forward health, all of AI, Babylon.

00:07:58: They all prioritize disruption over integration.

00:08:01: Segerberg van Roysen argued that AI won't revolutionize anything until we fix these structural issues.

00:08:07: Like siloed data and poor usability.

00:08:10: In a really unclear ROI, who actually pays for the efficiency gains?

00:08:15: That's a huge question.

00:08:16: These financial pressures are also changing how hospitals even think about equipment.

00:08:20: Kinesh Kamath described it as the uberola moment for healthcare.

00:08:24: The move to devices as service or DAT.

00:08:27: Right.

00:08:28: Why own a rapidly depreciating high-tech machine when you can just subscribe to the service it provides?

00:08:33: It's a massive shift from CAPEX to OPEX.

00:08:35: Ownership is becoming a liability.

00:08:36: And Paul Reckmans from Phillips noted that ninety-six percent of healthcare leaders are facing financial challenges.

00:08:41: Almost sixty percent have had to delay or cancel investments in new tech.

00:08:45: So the capital just isn't there.

00:08:47: Which means you have to fix the foundations first.

00:08:50: A CMK emphasized that open standards like FHIR are non-negotiable for AI to work at scale.

00:08:57: And we also saw some good news on the regulatory side in genomics.

00:09:00: We

00:09:00: did.

00:09:00: Lisa Gurry celebrated the FDA's new plausible mechanism pathway.

00:09:05: Which is designed to speed up access to personalized gene therapies.

00:09:08: That's a huge deal.

00:09:09: It shows policies trying to adapt to the science.

00:09:12: It does.

00:09:12: So when you look across all of this, from diagnostics to finance, what really stands out to me is that the tech that works is the tech that empowers the human.

00:09:21: It's not about replacing them.

00:09:23: It's about ambient scribes reducing burnout or surgical copilots enhancing precision.

00:09:28: It's all about augmentation.

00:09:29: Absolutely.

00:09:30: The technology that succeeds is the one that makes the clinician better at their core job, which is compassion and care.

00:09:37: If you enjoyed this deep dive, new episodes drop every two weeks.

00:09:40: 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.

00:09:49: And as a final thought, we have to look beyond all this shiny new tech.

00:09:54: Simon Philip Ross shared some OECD data that was, well, pretty striking.

00:09:59: What

00:09:59: was it?

00:09:59: Only three percent of all health spending goes into prevention.

00:10:03: and just fourteen percent into primary care.

00:10:05: We were overwhelmingly funding sickness.

00:10:08: Meanwhile, Vicky Britton noted that health literacy is stuck in the paper era.

00:10:13: Patients are navigating Reddit and TikTok for medical advice, sometimes out of sheer frustration with the system.

00:10:19: The real revolution won't just be an algorithm.

00:10:21: It'll be rewiring budgets to fund prevention and giving people the digital literacy they need to handle all this information responsibly.

00:10:27: Thank you for diving deep with us.

00:10:29: That is certainly something to think about.

00:10:31: We look forward to seeing you next time and don't forget to subscribe.