AI detecting diabetic eye disease
Artificial intelligence is being used in early screening for diabetic retinopathy, the most common form of diabetic eye disease, local medical experts said on Friday.
Diabetic retinopathy, or DR, can cause blindness if untreated and there is a rising number of people with the disease along with a rising number of diabetics.
DR has become the top cause of blindness in the working population, doctors said.
Early screening, proper diagnosis and intervention are effective for DR control and prevention and reduction of loss of vision.
Shanghai No. 6 People’s Hospital teamed up with Shanghai Jiao Tong University, Shanghai Key Laboratory of Diabetes and Shanghai Diabetes Clinical Medicine Center to develop an AI-based screening system for DR.
The DeepDR system can automatically read and evaluate fundus imaging and assist diagnosis.
So far, the system has been used in an epidemiological survey on diabetes complications covering 50,000 people across the country and a similar survey covering 20,000 people in Shanghai. It is also used by many hospitals to assist diagnosis.
Previous use found it can identify 86 percent of light diabetic retinopathy, which can be managed through proper control of blood sugar, blood pressure and cholesterol, doctors said.
A patient is screened diabetic retinopathy through an AI system at Shanghai No.6 People's Hospital on Friday.
