Ping An Insurance (Group) Company of China, Ltd. (hereafter "Ping An", the "Company" or the "Group", HKEX: 02318; SSE: 601318) announced today that five papers related to artificial intelligence (AI) technology in ophthalmic imaging by its smart city's smart health care team (Ping An Smart Healthcare) have been accepted by the top international medical imaging conference, the International Symposium on Biomedical Imaging (ISBI) 2020. ISBI is sponsored by the Institute of Electrical and Electronics Engineers (IEEE).

The Ping An papers cover various smart tasks related to the diagnosis of eye diseases, including quality control, screening, lesion detection, segmentation and quantization of retinal structure or lesion areas. The ISBI is one of the world's top academic conferences dedicated to mathematical, algorithmic and computational aspects of biological and biomedical imaging.

Ping An Smart Healthcare is affiliated with Ping An Smart City, a subsidiary of the Group. It uses the Group's three core technologies -- AI, blockchain and cloud computing -- to provide integrated diagnosis and treatment solutions. It includes elements such as smart triage and guidance, smart disease prevention, smart medical imaging screening/ diagnosis and smart clinical decision support.

In addition to the papers to be presented at ISBI, Ping An Smart Healthcare's AI Optical Coherence Tomography (OCT) screening system proposed in papers has completed clinical validation in numerous hospitals.

From 2018 to 2019, Ping An Smart Healthcare cooperated with Optovue, an internationally renowned OCT manufacturer, to develop multiple AI OCT products. The AI OCT screening system completed clinical validation in the Eye & ENT Hospital of Fudan University, Shanghai General Hospital and Shanghai Tenth People's Hospital at the beginning of 2019, and recently completed clinical validation in multiple prestigious hospitals, including Peking Union Medical College Hospital, the University of Hong Kong, S.A.R., China, Qingdao Eye Hospital and Shandong Eye Hospital.

The clinical validation of Peking Union Medical College Hospital covered image quality control, pathological changes screening, multiple lesions recognition and smart enhancement of OCT imaging. Among the data of over 900 eyes, the accuracy rate of the image quality control model, the pathological changes screening model and the lesions detection model were 99.0%, 96.8% and 97.5% respectively. Pathological changes screening can show whether there are abnormalities on images, while lesions detection can detect the locations of the pathological changes. The combination of both services can assist doctors in OCT image-reading. The smart enhancement of the imaging model saved an average of 46% of time and significantly improved the quality of imaging compared with traditional methods.

Professor Chen Youxin, ophthalmologist of Peking Union Medical College Hospital, said, "The utilization of AI technology may enhance the quality of OCT imaging, optimize the patient's experience in clinical examination and assist clinical doctors in image-reading, thus increasing work efficiency. Through technological innovation and model innovation, it may effectively solve the problem of insufficient medical resources in ophthalmology."

The clinical validation completed by Qingdao Eye Hospital and Shandong Eye Hospital verified the model function of recognition of anterior segment OCT multiple lesions and automatic corneal stratification. The accuracy rate of the multiple lesions recognition model on more than 1,700 verification images was 96.2%, which covered 16 types of common anterior segments' pathological changes. The automatic corneal stratification model can achieve automatic segmentation of the corneal epithelium and stroma and realize the calculation of the quantitative indicators such as corneal thickness and curvature, which may facilitate doctors' diagnoses. On more than 1,200 verification images, the doctors' overall recognition of the segmentation effect was 96.1%.

Professor Xie Lixin, ophthalmologist with Chinese Academy of Engineering and the President of Qingdao Eye Hospital, said, "New technologies such as AI are combined with ophthalmology clinical validation to realize early screening and standardized treatment of common diseases that lead to blindness such as corneal disease. It may effectively reduce the level of damage and the risk of blindness."

Dr. Xie Guotong, Chief Healthcare Scientist of Ping An Group said, "To date, Ping An Smart Healthcare has already provided smart ophthalmic imaging technology to more than 100 hospitals in 15 provinces and cities through Ping An Cloud, including Beijing, Shanghai, Shenzhen and Gansu. By deploying the AI model on the Ping An Cloud's platform, the hospitals can remotely transmit OCT images to Ping An Cloud after removing patient's information, and then send the model processing results back to the end users through the network."

Over the last few years, Ping An has constructed a health care ecosystem using leading AI technologies. Ping An Smart Healthcare's smart imaging system covers nine major human body systems with a model accuracy rate of more than 90%. It supports various pieces of equipment such as Computed Tomography (CT), X-ray, Magnetic Resonance Imaging (MRI), ultrasound, pathology, and the OCT fundus camera. In 2019, there were 120 million smart imaging pictures transmitted.

Ping An Smart Healthcare's smart imaging system has been applied at the country's top 10 key hospitals and key specialized hospitals, including Ruijin Hospital, affiliated with Shanghai Jiao Tong University School of Medicine, Huashan Hospital, affiliated to Fudan University, Fudan University Shanghai Cancer Center, Children's Hospital of Fudan University, and more than 3,000 primary hospitals in Guizhou, Hunan, Jiangxi, Guangxi and Shanxi.

Register as Visitor to CPhI China 2020!

-----------------------------------------------------------------------
Editor's Note:

En-CPhI.CN is a vertical B2B online trade platform serving the pharmaceutical industry,

for any copyright disputes involved in the reproduced articles,

please email: Julia.Zhang@ubmsinoexpo.com to motify or remove the content.