There has never been a more exciting time to be an optometrist. Artificial intelligence (AI), advanced imaging and increasingly sophisticated decision support tools are giving us access to information that previous generations could only have imagined. An interesting question is not “What can technology tell us?” but “What can we do with what it tells us?”

Our feature on Clinical courage explores the growing risk of algorithmic dependence. AI can flag abnormalities, analyse images and support referral decisions, but it cannot take responsibility for the patient in front of us. As Fiona Buckmaster MCOptom puts it: “AI cannot tell you the truth. It can only tell you a prediction of what it has determined is likely to be true.” That distinction matters.

Clinical judgement therefore becomes increasingly important, rather than less so. Kyla Black MCOptom describes it as a professional “spider sense” – recognising when something does not quite fit and having the confidence to investigate further. But judgement is not simply innate. It develops through knowledge, experience, reflection, feedback and education.

The same principle applies when technology gives us more things to see. 

Our feature on overdiagnosing examines how increasingly sensitive imaging can create clinical challenges as well as opportunities. Optical coherence tomography (OCT) has transformed eye care, allowing us to identify structural changes that may previously have gone undetected. But seeing more does not necessarily mean diagnosing more accurately. As College Clinical Adviser Daniel Hardiman-McCartney MBE FCOptom points out, OCT findings need to be interpreted alongside image quality, biological variation and the patient’s wider clinical picture. An AI image is a finding, not a diagnosis.

That makes the timing of the College’s new Clinical Learning in Practice approach particularly interesting. As the profession becomes more technologically sophisticated, future optometrists need to learn not only how to use new tools, but how to question them, interpret their outputs and integrate them into sound clinical reasoning.

And then there is the patient. Our feature on communication is a useful reminder that even the most sophisticated technology cannot replace a good conversation. Understanding why a patient has attended, how their symptoms affect their life and what concerns them provides clinical information that no scan can supply. As Dr Sandeep Dhallu MCOptom explains, getting to know the patient “allows you to understand what’s happening with their eyes and how it’s affecting their life”.

Perhaps that is the real challenge – and opportunity – facing modern optometry. We should embrace technology when it improves care, welcome imaging that helps us detect disease earlier and develop AI literacy rather than AI fear. 

But we also need education that prepares clinicians to question technology, interpret its outputs and apply them in the context of the individual patient.

The goal, however, is not to create better technicians. As James Wolffsohn FCOptom puts it: “AI will be part of everyday practice, but we’re training clinicians, not technicians.”

That distinction may prove more important than any algorithm.

Image credit | Caroline Andrieu