Uncategorized

How the NHS is preparing for supercomputers while still relying on fax machines.

How the NHS is preparing for supercomputers while still relying on fax machines.

In this Blog, we discuss how viable plans for AI and Machine learning are in the NHS and why old technology still reigns supreme.

In this one sentence, we could surmise both the current reliance on old technology and the fears and dreams of embracing the technology of the future. These were two headlines separated by just 4 days, on a national news outlet website. In a week where new doctors and pharmacists take to the wards and dispensaries of healthcare environments nationwide, the amalgamation of both cutting edge, and tried and tested technology is surprising as it is terrifying.

Fax machines, no longer classed as secure transmissions, are used to routinely send and receive personally identifiable information by healthcare providers. Widely used throughout the 80’s and 90’s, the practice still remains strong with many healthcare professionals accepting that they have no other choice to use the outdated communication method. Surprising when you think that since the invention of Fax machines, we have seen the fruition of technology such as SMS, email, Voice Over Internet Protocol (VOIP) and secure mobile instant messaging, technologies that have all evolved to take the place of the trusted Fax machine for everyday communication.

And yet, in the same sentence, Artificial Intelligence, a technology barely 10 years old, is being spoken about as a potential threat to not just the livelihoods of doctors, but to all skilled professionals. Surprising to think that in such a risk adverse environment as the NHS, the implementation of such new technology could be just years away, yet terrifying at the same time because as healthcare professionals, our jobs would change dramatically (note, I do not see technology and healthcare professionals as opposing forces, to create a truly world-class health service one will rely on another and vice versa).
  • Text Hover
Is technology ready to take our jobs and more importantly is the NHS ready to implement this new technology? 
Let's break this down. As healthcare professionals, we have learnt how to take in multiple points of information and use them in clearly defined ways to assess a characteristic of a patient. An example of this could be a patients eGFR or BMI. The result may itself then be used in another clearly defined pathway or be used as a marker for health on its own. We’ve learnt this over many years and had to fail, as well as succeed, in order to have the trust of the patients, the public and our employers. 

When you think about this as a function or a process, you can see that our roles are based around variables, rules and outcomes. Variables such as weight, blood pressure, mood, cholesterol. Rules such as NICE guidelines and outcomes such as the correct Antibiotics still work for the correct pathogen. In essence, a machine works in much the same way - it computes outcomes based on inputs and rules. It’s just that for the longest of time, the processing power available has not matched the needs of a system that is able to work as intelligently as you or me.

Two big things have changed, however, computational power has grown exponentially and it’s being used in much more creative ways. You may know the cloud as simply something that keeps your documents ready on all your devices, but it also enables the power of quite literally thousands of computers to be used in unison. To try and convey the impact of cloud computing in one sentence would do a disservice to its true capability, but for this argument, it is enough to say that cloud computing has made tackling the worlds biggest problems more accessible. You can rent time on cloud networks and save orders of magnitude in money vs building your own supercomputer. This means more people can not only afford to rent time, but they also have the budget to use the power in new ways.

Earlier in the post, I spoke about the time we have all taken to learn to get where we are. We have had to learn each thing we know. It sounds spectacularly obvious and is underlined even more when we talk about the experience that is required of a consultant or other high-grade position. Its the time they’ve spent honing their skills. With the advent of Artificial intelligence and Machine learning, this is no longer the status quo. 

Artificial intelligence allows machines to learn by themselves. The code that runs the computer is merely a teacher rather than a hardline colonel giving specific instruction. Pair this with the a computers ability to remember everything (minus hardware failures) and their amazing ability to use all of their power to focus on just calculation (as opposed to eating, breathing, wondering why time is moving so slowly, etc) and its clear that machines are able to learn much quicker than humans as it is. Now add thousands of computers connected to one big cloud, all simultaneously learning and sending their results throughout the network to improve it and you can see just how interesting AI is to health care policymakers. 

Like Content about healthcare and Tech?

We love writing about this stuff and if you like reading about it, you'll love being a part of our mailing list. Join by clicking the link!

So is the technology ready to take our jobs - No, the technology is mature enough to run an entire healthcare system. Google, IBM and even Amazon are players in this field so we know its not a shortfall of computing power that's holding us back. The coding and understanding of these networks still need vast improvements as well as the governance behind such systems.

Is the NHS ready to implement this new technology - Yes. The beauty of AI systems is that they are scalable. Do you need to add a new hospital to the system? Sure, just order more power from your tech provider. Worried that one system won’t talk to the other? No problem, the machine will figure out how to do this by itself. 

But perhaps the biggest reason that it’s a surprising yes to the last question is that new technology doesn’t only need to be better than what is replacing, it needs to be MUCH better to justify the costs for changing infrastructure and so forth. You’ve spent thousands of pounds over the years, upgrading your power lines, phone lines, buying new fax machines, computers with network access to printers so that you can communicate with any other person using the most common of identifiers - a phone number. No communication tool that has come out since has been as universally accessible and standardised as Fax. While all the other forms of communication may be better, none present as a far superior alternative when considering all factors.

AI, however, is different. yes, we are a far way off from understanding how far we can push the technology but already we can see that it will be far superior to anything that has come before it. And who knows, maybe even AI will end up using Fax machines too. 

Author


Avatar