Verily and Nikon to collaborate on machine learning in diabetes

While many were enjoying their end of year holidays, Google’s life sciences company Verily was busy completing one final business deal of 2016.

Verily’s alliance is with Nikon’s retinal imaging subsidiary Optos, which aims to leverage machine learning in the development of improved care solutions for diabetic retinopathy (DR) and diabetic macular oedema (DME).

The collaboration will look to create better screening methods for both conditions as well as assisted reading programmes to help specialists diagnose either disease.

Verily brings machine learning prowess to the deal whilst Optos will apply its core imaging technologies, including its retina imaging platform Optomap which can capture around 82% and 200^o of the retina in one image.

According to the World Health Organisation, the number of people with diabetes is rapidly increasing worldwide, particularly in low and middle income countries, leading to the WHO projecting diabetes to be the seventh leading cause of death worldwide by 2030.

Diabetes is a major cause of blindness; DR is caused by progressive damage to blood vessels of the retina, while DME (fluid accumulation in the eye often as a result of DR) is the leading cause of blindness in people with diabetes.

DR is often hard to spot in its earliest stages because of its asymptomatic nature, with loss of vision occurring only once the condition has reached severe stages.

In terms of machine learning, Google’s DeepMind Health has been busy establishing a relationship with the NHS, lending its tech to the Moorsfield Eye Hospital NHS Foundation Trust, the University College London Hospital, and to the Royal Free hospital.

In diabetes, the company’s machine learning capabilities have already impressed in DR and DME diagnosis, achieving an over 90% diagnosis accuracy for both conditions using archived images.

The new deal adds to Verily's multi-million pound joint venture with Sanofi, which aims to help people with diabetes by leveraging technology, medicine and professional care.