Artificial Intelligence in Ophthalmology: Replacement or Augmentation

Abstract

Artificial intelligence (AI) has speedily established itself at the forefront of modern ophthalmology practice. It has proven to be matchless when it comes to analyzing large data with uneasy patterns subsequently giving an insight.¹ But does artificial intelligence performs comparable to or even exceeds the clinician’s efficacy in completion of specific tasks? The answer to this question is not an easy one and will raise further questions as to whether artificial intelligence is able to replace ophthalmologists in clinical practice?

Artificial intelligence has demonstrated maximum utility in imaging based ophthalmic specialties. Its learning algorithms have proven to be very accurate in diagnosing and staging diabetic retinopathy, age related macular degeneration and even glaucoma with the aid of optical coherence tomography and fundus photography. However, in simple and clear cut tasks the AI softwares usually excel but when it comes to clinical correlation with a real time scenario AI is not free of limitations. This gap presents a challenge of correctly interpreting heterogeneous data and workflow integration.