Reality check needed before healthcare’s AI-centric future
AI can “bridge the gap between human and machine”, but ensuring solid digital foundations is imperative for long-term implementation. By Robert Barrie.
Credit: Jackie Niam via Shutterstock.
Artificial intelligence (AI) is booming in healthcare. From medtech startups to big pharma, it seems that if you are not using it, you will be left behind.
A 2023 report by GlobalData predicts that global revenue for AI platforms across healthcare will reach $18.8bn by 2027.
However, at the Med-Tech World 2023 conference in Valletta, Malta, a sentiment that peered through the glitz and glam of the AI transformation in health was the need for a solid digital infrastructure. Without appropriate foundations, the technology’s implementation could be hindered.
AI in medical imaging
Radiology is one of the fastest-growing areas of AI adoption in healthcare. The ability of automated software to reduce the workloads of radiologists and increase the throughput of scans means patients can be diagnosed quicker and with fewer errors.
Despite numerous companies deploying software-as-a-service for AI platform integration into imaging technology, Consultant radiologist and UK radiology tech company Hexarad CEO Dr Farzana Rahman said that healthcare is still quite far behind in technological foundations. Dr Farzana, speaking as part of a panel on the future of AI in medical imaging, said that although digital infrastructure is less attractive to discuss than AI, it is just as important.
She added that in some NHS hospitals in the UK, digital network infrastructure – even reliable wireless technology – sometimes comes up short. A robust digital framework is needed to ensure streamlined integration of AI into clinical workflows. If this foundation does not exist, then implementation of the technology – especially on the NHS – is harder to justify.
OneImaging co-founder and CEO Elan Adler and Modi Ventures general partner Sahir Ali both called on the need for data integration. Millions of scans and images are used to train algorithms for medical imaging software – pooling data could accelerate and standardise algorithm training.
The NHS digital landscape
Amid the latest AI-powered gizmos and gadgets on offer, the National Director for England’s National Health Service (NHS) Innovation Accelerator discussed the top priorities when deciding which technology to scale across the health service.
Konrad Dobschuetz, also chief enterprise officer at UCLPartners, spoke as part of a panel on the intelligent hospitals of the future.
“In my role, we look at innovations with a real-impact, real-life picture. We ask, what can impact the NHS positively and help right now? When we talk about the future, in the UK at least, it’s about how we maintain what the NHS is doing now for the next few decades,” Dobschuetz said.
“We’re not talking about the latest gimmicks; it’s about maintaining and improving the current standard of care.”
The NHS Innovation Accelerator has delivered around £38m ($46m) in savings by helping to scale high-impact innovations across the NHS via partnerships with 15 Academic Health Science Networks.
The UK Government’s stance on AI in life sciences is clear. Last month, the government announced a new £100m fund to help accelerate the use of AI in the sector.
However, despite ample funding and acceleration schemes, a lack of digital maturity and direction means it could be difficult to integrate AI into current systems.
“Many healthcare organisations are still maturing in their own digital strategy and architecture, which is the foundation needed to deploy new technologies like AI. Some healthcare organisations are still moving software from on-premises to the cloud for example, so there needs to be a maturity assessment,” said Patchwork Health COO and co-founder Dr Jing Ouyang, speaking to Medical Device Network.
“A related point is that regulations still need catching up, so IT teams are ill-equipped to fully assess new technologies and the risks/opportunities within these healthcare organisations.”
What about robotics?
Along with radiology, AI in robotics is growing at a considerable rate in healthcare. GlobalData predicts the robotic surgical systems market will reach $7.2bn in 2033, with a compound annual growth rate (CAGR) of 15.7%.
The market was catalysed by Intuitive Surgical da Vinci surgical system’s initial US Food and Drug Administration (FDA) approval in 1997, and a later expanded approval in 2000. Intuitive Surgical has the dominant share in the global robotic surgical systems market.
The investment landscape in healthcare is currently teeming with robotics-based companies. In 2019, J&J acquired surgical robotics firm Auris Health for $3.4bn. In October 2023, CMR Surgical raised $165m, bringing its total funding to more than $1bn and cementing its unicorn status.
Professor Manish Chand, professor of surgery at University College London, outlined how AI can bridge the gap between humans and machines. Unlike much of what grabs the public’s imagination, Professor Chand said the star of the show is the software, not the hardware.
From augmented reality, anatomical overlays, and sophisticated surgical navigation platforms, AI has become an essential fixture in the operating room.
“The fact remains surgeons are not 100% every single time. That’s how we need to use AI, so we can bridge the gap between human and machine. We can’t be arrogant enough to think we’re 100% every time we step into the operating room,” said Professor Chand.
But it’s not just about implementing the latest robotics with the most advanced capabilities, he explained.
“If you give a surgeon a toy, he or she will play with it. We need to make sure if we introduce technology, it solves a problem, it’s cost-effective, and it maintains the doctor-patient-relationship.”
The doctor-patient relationship was a recurring theme at the conference. Unlike the robots of sci-fi films, the ones used in healthcare are not autonomous. They are under the direct control of a surgeon. AI is, ultimately, an augmentative tool and should not act as a replacement for healthcare professionals.
AI reality check
“AI holds a huge amount of promise, but we should advise some caution. There is a lot of debate in the industry and unconscious biases and potential abuse of this technology – the risks are even higher in clinical settings where patient safety is at stake,” Dr Ouyang said.
There is no doubt AI will, however, become a permanent aspect of healthcare. In many fields of medicine, the AI revolution is already here.
“We have an ageing population with more complex healthcare needs and a global shortage of clinicians – AI can really augment clinicians and help them scale. That alone means it is worth investing in, but the regulations, safety, and the maturity of the healthcare organisations themselves need to be reviewed in this context,” Dr Ouyang added.
But a step back is perhaps needed, appreciating what is truly impactful for patients today, not tomorrow. After all, AI can only run after it has walked.