AI system able to detect eye problems as effectively as experts

New machine-learning system can detect more than 50 different eyes diseases with 94% accuracy


A new machine-learning system that is as effective as the best human experts at detecting eye problems and referring patients for treatment has been developed, according to The Guardian.

According to its developers, the new system is able to correctly refer patients with more than 50 different eye disease for further treatment with 94% accuracy, a standard akin to leading professionals in the field.

The technology has been developed by the AI-outfit DeepMind with Moorfields Eye Hospital NHS Foundation Trust in London and University College London.

"The number of eye scans we're performing is growing at a pace much faster than human experts are able to interpret them," said Dr Pearse Keane, a consultant ophthalmologist at Moorfields Eye Hospital. "The AI technology we're developing is designed to prioritize patients who need to be seen and treated urgently by a doctor or eye care professional."

Visit Innovation Enterprise's Big Data & Machine Learning for Pharma Europe 2018 Summit in Madrid on October 10, 2018

Keane added: "If we can diagnose and treat eye conditions early, it gives us the best chance of saving people's sight."

The use of AI in eye condition detection has led to the technology being able to take a more human-like and intelligible approach to analyzing complex optical coherence tomography (OCT) scans of patient retinas.

The technology has been designed to utilize five different machine-learning systems which have been trained using almost 900 clinical OCT scans, to first create maps of OCT scans. These were then analyzed by a further five machine-learning systems, which were trained on maps created from almost 15,000 OCT scans from more than 7,600 patients, which provided an individual referral decision.

These decisions will then be combined into one solid result, while any opposing or ambiguous results are reported to a human specialist.

The next stage of the initiative will see the AI system put through clinical trials and regulatory approval before it can be used for patient referrals. If granted approval, it will be available in all of Moorfields' hospitals for five years.

5 blockchain startups set to disrupt industry conventions small

Read next:

Five blockchain startups set to disrupt industry conventions