Inherited retinal diseases (IRDs), single-gene disorders affecting the retina, are very difficult to diagnose since they’re unusual and involve changes in one in every of many candidate genes. Outside specialist centers, there are few experts who’ve adequate knowledge of those diseases, and this makes it difficult for patients to access proper testing and diagnosis. But now, researchers from the UK and Germany have used artificial intelligence (AI) to develop a system that they consider will enable more widespread provision of testing, along with improved efficiency.

Dr Nikolas Pontikos, a bunch leader on the UCL Institute of Ophthalmology and Moorfields Eye Hospital, London, UK will tell the annual conference of the European Society of Human Genetics today (Saturday 10 June) about his team’s development of Eye2Gene, an AI system able to identifying the genetic explanation for IRDs from retinal scans. “Identifying the causative gene from a retinal scan is taken into account extremely difficult, even by experts. Nonetheless, the AI is capable of achieve this to the next level of accuracy than most human experts,” says Dr Pontikos.

The researchers were capable of utilize Moorfields Hospital’s vast database of knowledge on IRDs, covering over 30 years of research. Over 4000 patients have received a genetic diagnosis in addition to detailed retinal imaging at Moorfields, making it the biggest single-center dataset of patients with each retinal and genetic data.

Identification of the gene involved in a retinal disease is commonly guided through the use of the patient’s phenotype defined using the Human Phenotype Ontology (HPO). The HPO involves the usage of standardized and structured descriptions of medical terms of a patient’s phenotype, that are observable characteristics of a person resulting from the expression of genes, to permit scientists and doctors to speak more effectively. “Nonetheless, HPO terms are sometimes imperfect descriptions of retinal imaging phenotypes, and the promise of Eye2Gene is that’s can provide a much richer source of knowledge than HPO terms alone by working directly from the retinal imaging,” says Dr Pontikos.

The team benchmarked Eye2Gene on 130 IRD cases with a known gene diagnosis for which whole exome/genome, retinal scans, and detailed HPO descriptions were available, and compared their HPO gene scores with the Eye2Gene gene scores. They found Eye2Gene provided a rank for the right gene higher or equal to the HPO-only rating in over 70% of cases.

In the long run, Eye2Gene could possibly be easily incorporated into standard retinal examination, first as an assistant in specialist hospitals with the intention to get a second opinion, and eventually as a “synthetic expert” where such an individual shouldn’t be available. “Ideally, Eye2Gene software can be embedded into the retinal imaging device,” says Dr Pontikos.

Before its use becomes more widespread, the system might want to undergo regulatory approvals to exhibit safety and efficacy. This future use of AI has the potential to develop into a simpler, less invasive and more widely accessible approach to diagnosing patients, and to enhance their management and treatment. “We’d like further evaluation of Eye2Gene with the intention to assess its performance for several types of IRD patients from different ethnicities, several types of imaging devices, and in several types of settings, for instance primary vs secondary care. Clinical trials shall be required before our system will be deployed in clinics as medical device software,” says Dr Pontikos.

“Everyone knows that an image is price a thousand words, so we had some expectation that retinal scans interpreted by AI could out-perform HPO terms only. But we were still pleasantly surprised to see that, even when quite specific HPO terms were used, Eye2Gene could still do as well or higher than an HPO-only approach. We hope that AI will help patients and their families by making specialist care more efficient, accessible, and equitable,” he’ll conclude.

Professor Alexandre Reymond, chair of the conference, said: “While real-life experts are essential, the usage of AI will assist in mitigating biases and can allow diagnoses for all in the long run.”

Source:

European Society of Human Genetics