Ping An's AI-driven AskBob Doctor Earns Top Scores in Diabetes Management Competition
HONG KONG and SHANGHAI, Dec. 7, 2021 /PRNewswire/ -- Ping An Insurance (Group) Company of China, Ltd. (hereafter "Ping An" or the "Group", HKEx:2318; SSE:601318) announced that its AskBob Doctor, an artificial intelligence-driven consultation and treatment assistance tool, scored an average of 92.4 points, compared to a human team of six endocrinologists who scored 89.5 points in an International Human-Machine Competition on Diabetes Management. AskBob Doctor showed how it actively learns from clinical practice guidelines, the views of experts and a large database of case studies to provide effective diagnostic and treatment assistance to doctors.
The competition was jointly organized by the International Diabetes Federation (IDF), Ping An Smart City, Ping An Technology, Ping An Good Doctor, Peking University People's Hospital and Peking University International Hospital. It was part of the IDF Virtual Congress 2021.
Six doctors – practicing foreign and local medical professionals – from Peking University People's Hospital, Peking University International Hospital, Singapore General Hospital and Clinical Hospital of Ribeirao Preto (University of Sao Paulo) in Brazil took part in the event. AskBob Doctor was manually operated by a Ping An representative.
The competition event was hosted by Professor Luo Yingying and Dr. Zou Xiantong from Department of Endocrinology and Metabolism, Peking University People's Hospital. The judges included Prof. Jean Claude Mbanya, Honorary President of the International Diabetes Federation, Director of National Obesity Centre at the Central Hospital in Cameroon; Chen Liming, President of Tianjin Medical University Chu Hsien-I Memorial Hospital; and Dr. Bee Yong Mong, Head and Senior Consultant in the Department of Endocrinology at Singapore General Hospital. Andrew Boulton, President of the International Diabetes Federation and Professor at the University of Manchester, gave an opening speech at the virtual congress and the event was also attended by Professor Ji Linong, Director of Endocrinology and Metabolism at the Peking University People's Hospital and Xie Guotong, Chief Healthcare Scientist at Ping An Group and Deputy General Manager at Ping An Technology.
For the competition, the organizer provided nine cases of diabetes mellitus and its complications, including patient history and diagnostic findings, for a variety of diabetes scenarios with complications such as diabetic nephropathy, diabetic retinopathy, hypertension, hyperlipidemia and atherosclerotic cardiovascular disease. Competitors were provided medical records and could request further information on the patient's medical history, medical examinations or test results. The participants had to provide a final diagnosis and treatment solution, including primary and secondary diagnosis, treatment solution, lifestyle tips and follow-up procedures. The judges assessed the accuracy of the diagnosis, the rationality of the treatment recommendations, the comprehensiveness of other suggestions, and how it supplemented the necessary medical history and examinations.
The judges assessed the submissions provided by AskBob Doctor and the team of endocrinologists in an anonymous review process and scored the submissions based on diagnosis, treatment and other items. After a thorough evaluation by the judges, AskBob Doctor scored 92.4 points on average while the endocrinologists' team scored 89.5 points. The judges stated that the plan suggested by AskBob Doctor was more comprehensive and holistic. In addition, AskBob Doctor's response speed was impressive: AskBob Doctor took just a few seconds to complete one case, while a human doctor took nearly 20 minutes.
After the competition, Dr. Xie and Dr. Chaitanya Mamillapalli, an endocrinologist in Illinois (USA), delivered keynote speeches on the application of AI in the field of endocrinology and metabolism, followed by a panel discussion on the development prospects of AI in diabetes management joined by Professor Ji and world-known experts.
Professor Ji said, "This competition is a milestone event which would mark a new journey for conquering diabetes. Our team in Peking University people's hospital has been working on clinical decision support field for diabetes in the past 12 years with lots of experience. In recent years by collaborating with Ping An Group on advanced technology of artificial intelligence and big data, we can better empower doctors and improve the quality of medical services by leveraging a large amount of data, clinical guidelines, medical literatures and excellent treatment plans. As Professor Boulton had stated in his opening remarks, AI has a promising future. Our doctors will need to have more practice with AskBob in the future, and then will bring the more comprehensive, more individualized, and more importantly more warm diabetes care to our dear patients. "As a leading institution in the clinical and research field of endocrine and metabolic diseases, Peking University People's Hospital has always embraced innovative technologies and applied emerging medical intelligence technologies to the personalized diagnosis and treatment of diseases.
Mr. Xie said, "AI is meant to help doctors improve the quality and efficiency of diagnosis and treatment rather than replace doctors. Based on 40 million pieces of medical literatures, 200,000 drug instructions and 20,000 clinical guidelines, Ping An has constructed the world's largest bilingual medical knowledge graph and deep learning models, AskBob Doctor, which is expected to help doctors in diagnosis and treatment enhancing decision-making process by providing doctors with personalized and precise recommendations."AskBob Doctor has served more than 1 million doctors in more than 40,000 hospitals, providing more than 200,000 recommendations every day to help doctors to make better diagnoses and treatments.
Watch the competition video here.
– End –
For more information, please visit www.group.pingan.com and follow us on LinkedIn - PING AN.
Share this article