A few days ago, the first Diabetes Artificial Intelligence Competition was held in Wuxi. The Diabetes AI “Rining Sugar†and the Community Health Service Center doctors collaborated to “combat†the doctors from the tertiary hospitals. The two groups of doctors diagnosed the same diabetes cases and gave recommendations for medications, which were scored by a team of experts. The contest not only shows the latest developments in the application of artificial intelligence in the field of chronic diseases, but also explores the more possibilities for the future development of medical AI.
In the past two years, artificial intelligence (AI) technology has flourished and its application in medical treatment is in full swing. At present, the accuracy of morphological recognition such as imaging and pathology is even higher than that of doctors. However, there are not many AIs in terms of disease diagnosis and drug selection that require logical judgment. Currently, only the Watson oncologist developed by IBM is relatively successful.
In China, metabolic diseases are high, and the number of adults with type 2 diabetes alone has reached 114 million. In contrast, the current diagnosis and treatment of metabolic diseases in China mainly relies on tertiary hospitals, and the primary medical resources are seriously insufficient to meet the needs of patients for standardized detection and treatment and post-hospital health management.
How to prevent and control must have new ideas, new methods and new measures. With the development of the Internet and Internet of Things technologies, the idea of ​​AI for diabetes prevention and treatment has begun to emerge. Can there be an AI-assisted diagnosis and treatment system that helps doctors diagnose and treat? I believe this is the demand of many doctors. Therefore, the first artificial diabetes auxiliary drug engine for diabetes chronic disease, which was co-incubated by the Ningguang academician and his team, was born.
Academician Ning Guang served as Chairman of the First Diabetes Artificial Intelligence Competition
The first diabetes AI "Rui Ning Sugar" came to the fore
The form of the Diabetes Artificial Intelligence Competition is to invite five doctors from the tertiary hospital and the community health service center to participate in the diagnosis and to give a diagnosis of the five selected diabetes cases and give advice on the medication plan. Judging, the doctors at the Community Health Service Center cooperated with “Rui Ning Sugar†to complete the diagnosis and treatment plan.
The medications of the two groups of doctors will be scored by well-known experts in the discipline, a total of five judges. The scoring indicators include drug selection, starting dose and usage, each with a score of 10 points.
In the end, the average score of the tertiary hospital group was 72.9 points, and the average score of the community hospital + Ruining sugar group was 69.3 points.
Professor Ji Qiuhe, Xijing Hospital, Affiliated to Air Force Military Medical University (Figure 1)
Professor Peng Yongde, First Affiliated Hospital of Shanghai Jiaotong University (Figure 2)
Professor He Lanjie from Qilu Hospital of Shandong University (Qingdao Campus) (Figure 3)
Professor Shen Jie, the Third Affiliated Hospital of Southern Medical University (Figure 4)
In evaluating the results of the contest, Academician Ning Guang said: "I know that this result should be like this. This result shows that there is still a gap in the ability of community hospital doctors and tertiary hospital doctors in the diagnosis and treatment of diabetes, but AI can help community hospital doctors improve their ability. And level, narrow the gap." Especially in the context of graded diagnosis and treatment, two-way referral, only the level of doctors in community hospitals can be improved, so that the grassroots can be accommodated and patients can be kept.
Academician Ning Guang pointed out that although there are many applications of AI in the medical field, there are few focusing on the field of diabetes. The reason is that the diagnosis and treatment of diabetes is relatively complicated. The medications, dosages, etc. vary from person to person. It is difficult to breed high-level diabetes AI by dogmatically following the guidelines. From this perspective, Ruining sugar needs a lot of practice.
The iteration of medical artificial intelligence is very fast. Academician Ningguang said to the doctor on the spot that even if he can learn without eating or drinking, he can't keep up with the learning speed of AI. He is very confident that in two or three months, the ability of Ruining to help sugar will rise again, and it is still unclear when it reaches or exceeds the level of doctors in tertiary hospitals.
The artificial intelligence doctor “Rining Sugarâ€, which emerged in this man-machine competition, is the first artificial intelligence auxiliary drug engine developed by the National Metabolic Disease Clinical Medical Research Center and Ali Health.
It should be known that an endocrinologist must reach the level of “expertâ€, often after more than ten years or even decades of professional study, clinical practice, and advanced training, and in May this year, he was officially “employedâ€. "There has been a level that is indistinguishable from these experts. What is the secret behind it?"
The first diabetes artificial intelligence contest site
Knowledge model "Haina Baichuan"
Different from traditional artificial learning methods, Ruining Sugar has built a unique medical knowledge map, which integrates the latest diabetes guides and important medical literatures at home and abroad to realize real-time search and accumulation of subject knowledge. At the same time, AI experts use special algorithms to help build the connection between medical knowledge points, so that the machine can continue to "read" the new published literature.
Empirical model restores "expert thinking"
Different from traditional artificial learning, peer communication and long-term clinical experience accumulation, Ruining Sugar collects nearly 300 endocrinologists from all over the country to participate in data annotation, and refers to thousands of treatment experiences of desensitized real electronic medical records. It is a young but Experienced "expert player".
Show your skills in actual combat
In practical applications, doctors do not have to be limited to personal experience and knowledge. After entering information about patients, medical history, and inspections, “Rining Sugar†first standardized and integrated its desensitization information, and then combined medical knowledge. The data of the map provides recommendations for medication through artificial intelligence to provide more standardized and personalized diagnosis and treatment for diabetic patients.
At present, “Rining Sugar†can provide 3 sets of personalized and targeted treatments for diabetic patients with different conditions. Through the fusion of the "Knowledge Model - Empirical Model" dual model, it allows clinicians to obtain endocrine expert-level assistance and guidance in the diagnosis and treatment decision-making, thereby improving the level and achieving standardized diabetes diagnosis and treatment. Through continuous AI learning, the accuracy of “Rining Sugar†will continue to improve.
Diabetes AI goes to the grassroots
A doctor from the People's Hospital of Longkou City, Shandong Province, told the Arterial Network that she thought it would be very meaningful to hold such an event. Not only can it convey the latest developments in the field of medical AI, but it can also help doctors improve their diabetes diagnosis and treatment capabilities. Especially for the grassroots doctors, because of the difference in professional level and learning conditions, there is indeed a deficiency in the diagnosis and treatment of diabetes. With the help of medical AI, it is possible to continuously learn and improve.
In fact, this is also the idea of ​​the Ningguang academician and his team. As a clinically available auxiliary diagnosis and treatment system, “Rining Sugar†will enter the MMC National Standardized Metabolic Disease Management Center for trial.
As a co-developer of Ruining Sugar, Ke Yan, senior vice president of Ali Health, also pointed out that Ali Health has always believed that medical AI, which combines artificial intelligence technology and expert wisdom, can help grassroots doctors make more scientific and efficient. Decision-making, and constantly acquire the most cutting-edge medical knowledge, improve their medical skills, so as to benefit more patients, which is the initial heart of Ali Health Development Medical AI.
The National Standardized Metabolic Disease Management Center MMC is led by Academician Ning Guang, co-sponsored by the Chinese Medical Doctor Association, the National Metabolic Disease Clinical Medical Research Center, and the Shanghai Institute of Endocrine and Metabolic Diseases. It is supported by many companies such as AstraZeneca and Zhizhong Medical. With the philosophy of “one center, one standard, one-stop serviceâ€, we are committed to bringing personalized, high-standard disease management to different patients.
The continuously promoted National Standardized Metabolic Disease Management Center project is making great changes in the diagnosis and treatment of metabolic diseases in China: through one-stop medical treatment services, patients' medical treatment time is greatly saved, and data interconnection and artificial intelligence technology are used to create individuality. The high-standard, standardized, and easy-to-use screening, diagnosis, treatment, and follow-up procedures allow patients to receive expert-level treatment at community hospitals at their doorstep.
With the long-term linkage with tertiary hospitals and artificial intelligence-assisted diagnosis and treatment, it is expected to significantly improve the diagnostic rate and disease management level of primary medical institutions, and help improve the overall prevention and treatment level of metabolic diseases and the quality of medical services in China.
The linkage of tertiary hospitals with primary medical institutions and massive patient data will also promote the improvement of patient referral rates, provide feasible solutions for the promotion of chronic disease classification and treatment in China, and provide strong support for the formulation and optimization of relevant public health policies.
Academician Ningguang said: "At present, the construction of the center has already begun to show results. Next, we will vigorously promote MMC throughout the country, and strive to achieve the goal of 'establishing 1,000 MMC centers and managing 10 million diabetic patients' at an early date, and strive to After 10 years, the incidence of diabetes in China will be reduced by 1%, and all complications will be reduced by 10%, which will benefit more patients with metabolic diseases."
Fingerprint unlocking, faster, one-key opening
Micro slit cutting technology, high safety performance, beautiful atmosphere
Intelligent alarm system to keep track of property safety
Hotel Fingerprint Safe,Hotel Intelligent Safe,Jewelry Safety Deposit Box,Hotel Box Fingerprint Safe
Ningbo Reliance Security Technology CO.,Ltd , https://www.reliancesafes.com