Shaping the future of healthcare: Data, AI and Us

8 May 2024

Mike Beck, our Head of Electronics spent some time in Boston at the Device Talks Conference last week. Here are his take aways.

After last week’s DeviceTalks conference in Boston, which saw 110+ speakers from 80+ companies share their insights with over 1100 of us, I thought I’d take a few minutes to talk about where medical technology is heading, and what it means for us as designers, providers, and patients.

 

Mike with DeviceTalks attendees

Data – not just an android:

It looks like data is going to form the backbone of future healthcare systems. By leveraging new technologies and the insights they afford healthcare professionals, we can accelerate workflows and improve patient outcomes.

The increasing prevalence of health-trackers and smart-implants will provide doctors with more information about the patient’s condition. Inside the hospital we now have the opportunity for continuous patient monitoring, reducing the time that patients can deteriorate un-noticed and improve patient outcomes. This can be extended to outside the hospital too, allowing the patient to be ‘nudged’ into better lifestyle choices, minimising their need for clinical intervention in the first place.

Key players in this space: GE Healthcare, Pyrames, Bloomlife, Caretaker Medical

AI – a permanent resident?

Too much data might be just as problematic as too little. How do you draw out the useful insights for people under immense time-pressure?

It looks like AI might make itself indispensable here – taking the large swathes of data these devices generate and turning them into actionable information. We’ve already seen AI used in radiology to help analyse images for diagnostics, and there are plans for it to be involved in planning robotic surgery – creating ‘no-fly’ zones to minimise damage to the surrounding tissues.

Advancements in large language models, such as ChatGPT, might also make their way into surgical stations – helping medical staff write notes by dictation, saving them precious time to interact with patients instead.

Whilst all of this is fantastic for the healthcare professional and patient, we need to make sure that the AI is doing what it should – so that means lots and lots (and lots) of training, tuning and testing. The MHRA and FDA have jointly provided guidance on the development of Predetermined Change Control Plans (PCCPs). Historically the algorithm in your medical device has been locked as it goes to market – defeating much of the point of AI – but the existence of PCCPs allows manufacturers to iteratively update their algorithms once on the market,  ensuring safe operation of these devices over the device’s total product lifecycle and, hopefully, further improving patient outcomes. This means that the training, tuning and testing won’t stop when a product enters the market – and the development cost might just continue to ratchet up whilst the unit is on sale.

Key player: Kebormed

And what about us?

It’s no secret that the majority of medical devices (and drugs) have been designed by white men for white men, and indeed tested predominantly on white, middle-aged men. So, good news if I need medical treatment in my 40s – all these devices have been optimised for me! But what about women, people of colour, children and the elderly? These devices cannot reasonably be a ‘one-size-fits-all’ solution.

When we couple this obvious issue, with the reduced access to (accurate) healthcare provision afforded to these groups, we see a huge disparity in global healthcare equity. Whilst companies such as Medtronic are already making concerted efforts to address this, it falls on every single one of us to be more intentional in our decision-making. Only then will we ensure that we don’t just design for (and test on) the status-quo, and actually include everyone – regardless of sex, body-shape, age, ethnicity and (dis)ability.

Key resources: Go Red for Women | The American Heart Association’s signature women’s initiative,

Looking ahead – the challenges and opportunities

Increased data and the inclusion of AI in new medical products opens up several avenues to us, as designers, to improve the patient experience and success of any medical treatment. Decentralised healthcare systems will reduce the barriers for patients to access the treatment they need, whilst additional information available to healthcare providers will allow them to better diagnose illnesses and plan any interventional surgery.

Before this becomes a reality, there will be technical and regulatory challenges to solve – such as how to ensure data security and validity across the full device lifetime – but with a concerted team effort from designers and the regulatory bodies, I see no reason why these can’t be solved. Instead, one of the key challenges will probably be money. How do we ensure that this healthcare becomes available to those who need it, and not just those who can pay or fit into the most profitable sector?

If we can solve these though, the future of global healthcare provision is promising.

 

 “We engineer the extraordinary” – Rajit Kamal, VP of Robotic Surgical Technologies (Medtronic)