Radhika Rampat, MBBS, FRCOphth, on AI's growing role in IOL selection, keratoconus screening and anterior segment imaging

At the 44th Congress of the European Society of Cataract and Refractive Surgeons (ESCRS), held in London, Radhika Rampat, MBBS, FRCOphth, discussed the growing role of artificial intelligence (AI) across cataract and anterior segment ophthalmology—from lens power calculations to corneal disease detection and future patient-facing tools.

Critical appraisal before adoption

Rampat noted that many clinicians approach AI from the standpoint of a layperson, without fully understanding what the term means. For several years, she has given talks on how to critically appraise ophthalmology journal articles on AI, arguing that clinicians need to understand a paper before they can adopt its findings into practice.

AI in IOL selection and keratoconus screening

In her own practice, Rampat already uses 2 AI-supported tools. The first is the ESCRS online calculator, which incorporates a number of AI-based intraocular lens (IOL) power formulas. The second is the MS-39 device from CSO, which has AI capability built in to help indicate whether a patient has keratoconus. She emphasised that clinical judgement remains essential alongside these tools.

Standardising anterior segment imaging

Rampat has also published research on AI in the anterior segment, an area she said is less studied than the retina despite its many potential applications. The main barrier, she explained, is standardisation of imaging: photographs of the front surface of the eye are currently captured with less consistency than retinal images. Newer devices, such as Light Field, are beginning to offer standardised slit-lamp measurements, which she believes could be a game-changer for building the data sets needed to establish diagnostic baselines and, for example, recognise conditions such as Fuchs' endothelial dystrophy from a slit-lamp photo or video, regardless of where a patient is being seen.

What's next: patient-facing AI

Looking ahead, Rampat pointed to the pace of change in AI adoption, noting that predictions about slower uptake have consistently been overtaken by faster progress. She highlighted emerging tools that would let patients discuss upcoming surgery directly with a language model populated with their relevant information, describing the current moment as “really exciting times” for AI in ophthalmology.