“Where the responsibility lies is a key question that needs to be answered though these developments”

AOP outgoing chair, Professor Julie-Anne Little, on technology evolution and the role of artificial intelligence and machine learning in optometry in the future

An animation of three human hands and one robot hand placing their hands together in a circle.

I joined the General Optical Council as a qualified optometrist in 2003. Back then, while fundus photography was available, it wasn’t embedded in practice in the way that it is now. I worked in the hospital environment around that time and had also seen some interesting technology in regards to glaucoma monitoring. At the same time, the likes of the IOLmaster was relatively new for ocular biometry ¬and there would have been a transition over to this instrument from working with more traditional contact biometry techniques. That was the type of tech available, and everything was absolutely still on paper.

Over the past two decades, we have seen a massive shift in electronic care records affecting how we work and what we can do in terms of our ability to bring everything together digitally on a patient record – from patient notes, to scans and referrals written electronically. Yet with developments to these electronic systems, even 25 years later we’re still not connected well between primary and secondary care – it’s very much a patchwork across the postcodes of how well we can connect with both NHS colleagues when we refer, but also how well we can see general electronic case record information.

The arrival of AI

Artificial intelligence (AI) is here. Machine learning assisted information is happening – it’s already what we are using in many applications on our smartphones including and in geographic information systems.

In terms of healthcare, research trials and outcomes continue to be shared at pace – there’s a lot of research into wearable digital tech, as well home monitoring, for example, but clinicians still play a key role in how that data is interpreted to direct patient care.

The uncertainty isn’t exactly in where it might go, but in the regulation that is needed around it; the ethical, regulatory and legal frameworks. It feels like this is really overdue now


There are a lot of positives that we, as a profession, could harness with AI. There is a lot of data that AI can efficiently sift out, while flagging the stuff of interest. However, for me, the uncertainty isn’t exactly in where it might go, but in the regulation that is needed around it: the ethical, regulatory and legal frameworks. It feels like this is really overdue now.

I think there is a lot of fear in the profession in relation to AI, and a sense of ‘Is this going to replace us? Is everything going to become automated? And is the clinician no longer going to be needed?’ This is natural, and while I think part of the clinical processes will become automated, I also feel there will always be a need for the interpretation and understanding of that information, and we need to be custodians and fully engaged in the quality of the information going into AI/machine learning algorithms to make sure they are answering our questions adequately.

Where the responsibility lies is a key question that needs to be answered through these developments. Is it the company who built the machine that will be responsible for the decision-making, or is it the clinician who is deploying it and using that decision-making process? That is a real concern in terms of medical malpractice and clinical negligence.

The patient in focus

In terms of AI and patients, I’m certain in the future there will be developments in home monitoring, and this insight this will give into daily life will become more and more useful. And while it’s hard for me to imagine somebody being able to take their intraocular pressure at home at the moment, that certainly might be possible in the future.

Patients could benefit greatly from developments in home monitoring tools that, for example, in the future might be able to flag when somebody is showing a sign of an issue and sending this forward to be seen.

This type of home monitoring is not a pipe dream, it is already in existence. For example, I have recently heard about a project using optical coherence tomography for at home monitoring of AMD. The premise saw people at risk of wet AMD scan their eyes in a home-based OCT device every day with the data fed through and flagged if there signs of wet AMD, which was then signposted for prompt treatment. This has positives for the efficiency of these types of clinics, ensuring that the people who need to be seen are seen in a timely manner.

The ability of this type of monitoring to stratify risk and urgency is really valuable, and as optometrists in primary care I think we could be a real part of that process too. We could be the community connection hub where patients go to have images taken. This again highlights the importance of having IT connectivity in place for all because there will be touch points that need to be involved in any kind of machine learning or AI.


Overall, it strikes me that whilst there’s a rise in technology and its ability, there’s also a rise in the public really valuing human contact. People remain focused on availability and shopping locally, and post-pandemic people place a real hunger and value on human contact and expertise. I am confident that the profession is and will continue to be well-placed to provide that in the future.

About the author

Professor Julie-Anne Little is an optometrist, outgoing chairman of the AOP Board, and a senior lecturer at Ulster University

• As told to Emily McCormick.