Recently, the first computer vision summit jointly organized by Tencent's three AI laboratories and one of the official journals of the American Science Association's "Science" series was held in Shanghai. This is the first time that U-Lab has taken the initiative to publicly appear. It is also the first press conference sponsored by the technical team. At the same time, U-Lab is also officially upgraded to Computer Vision Lab.
The special nature of this meeting implies Tencent's determination to accelerate the layout of the AI ​​track. Although Tencent has made great achievements in the AI ​​field, Uto has delivered outstanding products in the fields of retail, social, industrial, entertainment and medical , but companies can't just look directly at the present. Tang Daosheng, senior executive vice president of Tencent, said at the meeting: "In the future research and development, the new laboratory will not set KPI. I hope that in the future, the machine will not only be 'visible' but also 'understandable' Even in the future, it can help humans to 'see further.' This is a very challenging basic research work, and we will have patience and long-term continuous investment in the future."
In the second half of the meeting, Professor Qiao Youlin, Professor of Epidemiology, Cancer Hospital, Chinese Academy of Medical Sciences, member of the Chinese Medical Association Ophthalmology Branch, Shantou University, Hong Kong Chinese University, and Shantou International Eye Center Professor Yan Lingping, Imperial College of Technology, London, Hamlyn Medical Director of the Robotic Robot Center, co-founder, Yang Guangzhong, editor of Science and Robotics, Dr. Qian Tianyi, Director of Tencent Medical Health Technology Committee, Dr. Zheng Yefeng, the head of medical AI of Tencent Youtu Laboratory, and other top "brains", on AI medicine The future will be brainstorming and talk about the current and future of AI medicine.
Three directions and three challenges of “AI+Medicalâ€
Yang Guangzhong pointed out that the application of AI in medical field can be summarized into three development directions: one is the optimization of medical procedures; the second is the diagnosis and prediction of diseases, which is the AI ​​medical image processing focused on most AI medical programs today; Precision medicine , especially precision surgery. "The lesions that are now discovered are getting earlier and smaller, and it is difficult to pinpoint the eye or traditional surgery."
As a benchmark for domestic AI medicine, Tencent Yingying has realized the use of AI medical image analysis to assist doctors in screening for esophageal cancer, pulmonary nodules, diabetic retinopathy, colorectal cancer, breast cancer and other diseases, as well as assisting doctors with AI auxiliary engine. Identify and predict more than 700 disease risks and cooperate with more than 100 top three hospitals in China. The rapid development path of Tencent's film reflects the development track of AI medicine.
According to Dr. Qian Tianyi, Director of Tencent Medical Health Technology Committee, from the technical point of view, medical AI may play a greater role in the three types of medical fields in the future: the first category does not require high years of experience and rich experience, but requires a large number of With the energy to judge the scene, AI can achieve very good results, save the doctor's energy, improve the diagnostic efficiency, and get the clinician's working time out of trivial and small things. The second category is the primary health care system lacking specialists. The training of specialists takes several years, requiring time and scenes and teachers. By learning the knowledge of specialists, AI products can become assistants to grassroots doctors in the absence of specialists and provide them with specialist knowledge. The third category is precision medicine. AI can not only find the lesions, but also measure the accurate density and volume of the lesions. This is more accurate and less error-free than manual measurement by doctors. The AI ​​assistant extracts accurate data to the doctor, allowing the doctor to save time with tools, shifting energy to breakthroughs in scientific research, and making medicine move forward.
"China's medical resources are unevenly distributed, and we hope to provide better medical conditions for the people through the development of medical AI." Dr. Qian Tianyi also pointed out that although "AI+Medical" is an opportunity, it will also face many challenges: for effective application. The data of AI training requires a lot of manpower, material resources and financial resources; more cross-border and interdisciplinary talents are needed. Both technical and medical knowledge can be combined to integrate technology accumulation and clinical application scenarios. It is more suitable for medical AI products used by doctors; in clinical registration, it is also necessary for competent authorities and industries to explore standards for evaluation and evaluation of AI medical products.
Multi-species research see Tencent AI development
"AI+Medical" is a brand new "health and labor combination" field. Driven by technologies such as big data, artificial intelligence, and cloud computing, the growth of the "AI Medical Assistant" has filled the medical community with expectations.
1. Cervical cancer screening
Qiao Youlin, a professor at the Epidemiology Laboratory of the Cancer Hospital of the Chinese Academy of Medical Sciences, successfully developed the HPV virus screening method, which directly contributed to the birth of the HPV vaccine and vigorously promoted the popularization of cervical cancer screening in China. In his view, the development and application of artificial intelligence electronic colposcopy in China not only improves the speed and accuracy of screening, but also provides technical support for remote screening.
According to WHO, by 2030, 90% of girls under the age of 15 will have HPV vaccine, and 70% of women over 35 will have at least two HPV-led screenings. “After two screenings, the electronic colposcope came in handy. I especially look forward to the day when our AI electronic colposcopy products come out, where other developing countries around the world do not have such good medical services, as long as there is a simple Electronic colposcopy equipment, the grassroots doctor through AI remote assistance, dozens of seconds to know where to sample, where the lesions."
It is reported that under the guidance of Professor Qiao Youlin, Tencent Yingying and Youtu Lab are promoting the research of early screening AI system for cervical cancer, creating a screening test tool for cervical cancer intelligent screening, assisting gynecologists to quickly identify cervical precancerous lesions and cancer. The lesions are fixed in the area to develop a corresponding biopsy and treatment plan.
2. Screening for sugar net lesions
In the field of ophthalmology, the application of artificial intelligence also allows doctors to see the prospects for innovation. Shantou University? The Chinese University of Hong Kong and the professor of Shantou International Eye Center, Yan Lingping, said that artificial intelligence is the most widely used and most studied in the field of ophthalmology. Diabetic retinopathy. "We hope to develop an artificial intelligence that can recognize all fundus diseases." product."
In October last year, the Western Ophthalmology Alliance teamed up with Tencent to build a joint laboratory for artificial intelligence medical imaging. The two sides will rely on the leading artificial intelligence technologies such as image recognition and deep learning of Tencent to conduct in-depth research and practice on fundus screening. Assist doctors to perform fundus screening, effectively improve the accuracy of diagnosis, and make up for the gap between medical resources in the western region and the uneven distribution of medical care.
3. Lung cancer screening
Dr. Zheng Yefeng, the head of the medical AI of Tencent U-Lab, used AI-assisted lung cancer screening as an example to introduce the rapid advancement of AI medical technology: In 2014, the FDA approved the CT product of lung cancer, when the traditional machine learning method was accurate. The rate is not high, 90% sensitivity, there may be more than 6-7 false positives in a CT scan; after deep learning technology, you can do better, you can increase the sensitivity by more than 95%, and make false positives. Drop to 1-2 or lower. He pointed out that with the rapid development of AI technology, China's AI medical applications have the opportunity to achieve leapfrog development such as high-speed rail and mobile payment. “We can work with hospital experts to train an AI algorithm that can approach or reach their level and finally deploy to the grassroots. We can really improve the accuracy of primary care with new technology.â€
The future development of AI+ medical
When talking about the core competitiveness of Tencent Youtu in the field of medical AI, Dr. Zheng Yefeng told the arterial network reporter that the current AI+ imaging industry is very competitive. Before they passed the approval of NMPA (CFDA), everyone talked about themselves. The results of the data are difficult to judge. Nevertheless, as lung cancer with high morbidity and mortality, we must continue to invest in ensuring its research and development to ensure the integrity of the product to promote the overall strategic layout. â€
As the "national team" of artificial intelligence, Tencent's strategy of comprehensive layout continues to advance, and is not limited to products. The cooperation with Science magazine at this conference highlights Tencent's emphasis on science and technology culture. The “Uniform Science†applet displayed at the conference can also easily allow C-end users to enter the medical ecosystem.
At present, Tencent's products have already cooperated with more than 100 domestic hospitals. This is a unique advantage and a long-term investment. Under the overall layout, once the approval results are released, Tencent Yingying will occupy the highland in the future “AI+Medical†industry.
For the most common waxy and sweet corn on the market, waxy corn has a higher nutrient content than regular corn, containing 70-75% starch (and almost all straight-chain starch), more than 10% protein, 4-5% fat and 2% multivitamins, with more grains, VA, VB1 and VB2 in protein than rice, with the highest fat and VB2 content. Yellow maize also contains carotenoids, such as rice and wheat. The molecular weight of waxy maize starch is more than 10 times smaller than that of ordinary maize, and the starch makes glutinous rice sticky and soft, softer than ordinary hard maize. It is more than 20% more digestible to eat than regular maize and it is suitable for people with less than perfect teeth. At the same time, it is not suitable for diabetics because of the very high content of straight-chain starch (a polysaccharide).
Waxy maize is also known as sticky maize. The grain has coarse, waxy endosperm, similar to shiny, glassy (clear) grains such as hard and dent maize. Its chemical and physical characteristics are controlled by a recessive gene, which is located on chromosome 9. 100% of the starch in the endosperm is straight-chain starch.
Healthy Corn Food To Eats,Fresh Corn Kernel,Instant Food Corn,Organic Sweet Corn
Jilin Province Argricultural Sister-in-law Food Co., Ltd. , https://www.nongsaocorns.com