Yiwei Technology: years of precipitation and technology exploration to crack the AI ​​ophthalmology problem

There is a rumor in the "Yellow Emperor's Internal Classic" written by the traditional Chinese medicine practitioners in the Yellow Emperor: "The upper work is not cured, the middle work is cured, and the work is finished." However, due to the scarcity and imbalance of medical resources, disease prevention and early screening are often neglected, and resources are more invested in the treatment, making “medical illness” an unfulfilled “dream”.

The emergence of artificial intelligence makes the combination of human intelligence and machine efficiency organically, thus making the realization of the "medical disease" dream possible. The arterial network "2018 Medical Artificial Intelligence Report" data shows that lung nodule screening and sugar net screening have become the two most popular areas of medical artificial intelligence. In addition to the continuous investment of giants such as Tencent and Google, major startup companies Also, you earn me and boost the development of the industry.

After three years of development in 2016, 2017 and 2018, there is still a threshold for AI to be applied to ophthalmology: how to break through the bottleneck of AI technology and realize the commercialization of products? And establish unique core competitive advantages in the market and industry. To this end, the arterial network interviewed many years of experts in the field of artificial intelligence ophthalmology, the founder and CEO of Yiwei Technology - Ke Xin.

Do the real problem solving AI, not just just add

Compared with peers, Yiwei Technology is a rising star, but Ke Xin is a pioneer in this industry. He was one of the earliest pioneers in ophthalmic artificial intelligence. He graduated from the Chinese Academy of Sciences and immediately entered the Daheng Image R&D Center for intelligent ophthalmology. R&D, and then as the team leader, responsible for product development and promotion. Daheng Image is one of the earliest companies in the field of ophthalmology in China. Since 1993, it has been working with Beijing Tongren Hospital on the subject of ophthalmology of the Ministry of Science and Technology.

In 2015, when the domestic AI ophthalmology was just emerging, Ke Xin led the team and director Yang Jinkui of the Department of Endocrinology of Beijing Tongren Hospital to jointly develop the “Computer Vision-based Sugar Network Screening Intelligent Analysis System”, and then promoted to dozens of domestic top-level companies. Public hospitals and private hospitals are used. In 2018, in order to further realize the value concept of "the doctor is not sick", Ke Xin created the technology of Yiwei, and hopes to use the eye as the entrance to open the era of AI disease early screening, and promote the development of precision and personalized medicine.

Based on Kexin, the Yisui technology team is based on technology and products. The backbone members are all from the Chinese Academy of Sciences and Tsinghua University. Yiwei Technology attaches great importance to technology and believes that technology is the foundation of AI. Its unique combination of computer vision and deep learning technology provides efficient and effective screening for eye diseases and systemic chronic diseases through original technical routes. A low cost solution. At the beginning of Yiwei Technology, in the industry's highest specification "World Artificial Intelligence" technology competition, it broke through more than 40 peer companies and won the second place in the final. At present, Yiwei technology products have been implemented in many hospitals across the country, and have reached cooperation with Novartis, Pfizer and other enterprises.

From the perspective of industry demand, the entry of artificial intelligence, for the screening of fundus diseases is not a simple icing on the cake, but a solution to the need. Artificial intelligence will become a tool to reduce the burden on doctors and improve the efficiency of early screening. AI+ ophthalmology is also considered by industry professionals to be the most likely area for large-scale commercialization. However, it is undeniable that there are still many barriers and difficulties in AI, especially how to achieve breakthroughs in establishing links between artificial intelligence and doctors and patients. Compared with other companies in the same industry, Yiwei Technology has taken a different path in this respect.

On the underlying technical support, Yiwei Technology used computer vision and deep learning technology to extract fundus lesions and quantify the feature data. Then, using data mining and big data analysis methods, the fundus image was analyzed in depth, and finally the AI ​​fundus image assisted. Diagnostic system.

Ke Xin also introduced us to the AI ​​highlights of Yiwei Technology: "Our AI not only has conclusions, but also identifies key lesions, key structures and key locations, and also obtains quantitative data, which is in line with domestic counterparts. In other words, AI's AI can not only help doctors diagnose, but also clearly tell the doctor the basis and source of the diagnosis."

Using technology first-mover advantage to solve AI landing problems

According to the choice, the node enters the AI ​​fundus field at this time, not just because of the tuyere. Ke Xin believes that market education is basically completed, the market for ophthalmology + AI has gradually opened up, and it has begun to show more pain points and needs. At the same time, it has put forward higher requirements for technology and talents. This is a good opportunity. But if the product can't fall into reality and fall into homogenization competition, then meeting the industry's needs will not be mentioned. Based on this, Yisui Technology took the lead in the field of the industry that has not yet been involved in the industry through the accumulation and precipitation of the founding team for many years, avoiding the Red Sea market of sugar net screening, and uniquely analyzing the fundus images and mining more quantitative information. In order to achieve the analysis of multiple diseases in the fundus, it will provide sufficient data support for accurate medical and personalized medicine in the future.

"I am a AI technology person. I was responsible for the research and development of fundus images in Daheng. I have been doing image recognition for nearly 10 years. At least in China, I belong to the group of people who are the first to work in ophthalmology AI. One of the characteristics of our team is that we With a deep medical heritage, not a pure IT background, the founding team is an expert in the field of medical imaging, which ensures that we have unique advantages in technology research and development, so that our products can be realized faster. Clinical landing," said Ke Xin.

Ke Xin also said: "At present, everyone's understanding of AI images is still a bit simple. The industry has to achieve real landing, in fact, the threshold is very high. Technology, data, medical knowledge and landing scenes are the barriers of this industry, but the most The core barriers are still technology and talent. Artificial intelligence is not just a method of acquiring images and then training. It must consider full-stack AI, otherwise it will be lacking in some practical considerations. Just like many people know about development. The algorithmic engineer of Alpha Dog is a master of Go. Therefore, the medical AI wants to land, and the technicians themselves need to have a deeper understanding of medical images. And our product logic is the key to the use of computer vision by algorithm engineers. The elements are extracted and then analyzed. So our results are intuitive, interpretable, and quantifiable, so that the information mined by the AI ​​is transparent to both the doctor and the patient. At this point, the technology of domestic counterparts is the same."

At the end of the interview, Ke Xin also put forward the idea that is being constantly verified: “Many medical AI practitioners now have fewer medical backgrounds, which has caused one of the biggest problems in the industry, and the products are difficult to reach. For medical treatment, we It should be awe-inspiring, so that the product can really solve the problem, to help the doctor, instead of replacing the doctor. The doctor can not be replaced, but with AI, the doctor can return to his position better, this is the progress of the times! ”

Handpiece Lubricating Machine

Dental Handpiece Lubrication Machine,Dental Lubricating Machine,Handpiece Lubricating Machine,Dental Handpiece Lubricant Device

Foshan Ja Suo Medical Device Co., LTD , https://www.fjoralinstrument.com