On March 27, 2018, NVIDIA's GTC 2018 (2018 GPU Technology Conference) was held in San Jose, California. This multi-day conference has attracted a lot of attention from AI practitioners around the world. On the March 29th schedule, the representative of China Medical Artificial Intelligence Enterprise Airdoc gave a detailed introduction to the development and application of artificial intelligence technology in the medical field in a 25-minute speech. The ophthalmologist from China's top public hospitals took the case form. The audience described the real situation in which artificial intelligence plays a role in hospital clinical practice.
Airdoc artificial intelligence helps primary care
Professor Li You from the Department of Ophthalmology, Shanghai Changzheng Hospital attended the conference as an expert representative, and explained in detail the specific application of artificial intelligence in hospital clinical practice--how artificial intelligence technology can help doctors' daily work.
On March 27, 2018, NVIDIA's GTC 2018 was held in San Jose, Calif., and attracted extensive attention from AI practitioners everywhere. On the March 29th schedule, the representative of China Medical Artificial Intelligence Enterprise Airdoc gave a detailed introduction to the development and application of artificial intelligence technology in the medical field in a 25-minute speech.
China's medical resources are unevenly distributed. Most of the high-quality medical resources are concentrated in large cities and large hospitals. Many remote areas lack quality medical resources. Professor Li Yu has repeatedly followed the medical team's support activities into remote areas such as the plateau, Xinjiang, and Tibet. There are not only drugs and doctors, but many people have to go to hospital support points for a long time. In the Haibei Tibetan Autonomous Region of Ningxia, where Professor Li had been to, several women were blind and had never left the mountains for the rest of their lives. The staff of the local Disabled Persons' Federation spent three days riding them to pick up Professor Li for medical treatment, but because of their long time for diabetic retinopathy, they have lost the best treatment opportunity.
Traffic, environmental and economic problems have plagued many patients, and doctor support activities alone cannot solve problems effectively. In order to improve the medical environment at the grassroots level, Changzheng Hospital and Airdoc have jointly developed a chronic patient intelligence algorithm. Hospital experts label the lesions one by one on a large number of retinal images. Each map is cross-labeled by several ophthalmologists. The accuracy of the algorithm is comparable to that of a professional ophthalmologist at the top three specialist hospitals.
Professor Li Yu introduced: "Now the Long March Hospital has deployed Airdoc's algorithm in many grassroots hospitals. After the patient took a fundus photo in the primary hospital, he will soon receive an artificial intelligence interpretation and a doctor-confirmed screen on his mobile phone. Check the report - not only can you see the disease identification results, but also keep the retina photos in the future if necessary."
Zhejiang patient Mr. Sheng Lao is in the local county hospital cooperating with Changzheng Hospital. After the rapid identification of diabetic retinopathy by Airdoc artificial intelligence chronic disease identification system, he was promptly referred to Shanghai Changzheng Hospital for active intervention treatment. Now his vision has recovered. Very ideal.
Airdoc Artificial Intelligence Community Chronic Disease Screening
There are about 12 million sugar net patients in China, who may face the risk of blindness in the morning and evening. There are more than 100 million cataract patients and over 9.4 million glaucoma patients over 40 years old. These diseases cause one person to be blind every minute, putting a heavy burden on patients and families. The number of ophthalmologists is limited, and 20% of county-level hospitals cannot set up ophthalmology because there is no ophthalmologist. Blindness is not only a huge social problem, but also a serious medical problem.
In order to solve the problem of screening for chronic diseases at the grassroots level, Airdoc has collected millions of fundus photographs from a number of top hospitals at home and abroad, and asked several ophthalmologists to cross-label and select the common label as the gold standard. Building a multi-layer convolutional neural network based on a large amount of data, NVIDIA's powerful GPU provides powerful technical support for the development of algorithms in the field of medical artificial intelligence.
Airdoc's successful artificial intelligence chronic disease screening algorithm can automatically identify a variety of common fundus diseases and chronic diseases such as diabetes and hypertensive fundus diseases. On the main indicators such as sensitivity and specificity, the results are exactly the same as those of human doctors.
Puyang Linshi Medical Supplies Co., Ltd. , https://www.linshimedical.com