Is there any weakness in myopia prevention and control? Artificial intelligence testing "myopia prediction"
original title: is myopia prevention and control weak? Artificial intelligence testing "myopia prediction"
the report of the World Health Organization shows that the number of myopia patients in China has reached 600million, among which the myopia rate of teenagers has surpassed Japan, ranking first in the world, causing great concern from all walks of life. In addition to laser surgery, myopia cannot be cured, but can myopia be predicted in advance and personalized targeted intervention? AI has targeted this 100 billion level market. The reporter learned that at present, medical AI enterprises such as airdoc have laid out the myopia prevention and control market, and the "Ai ophthalmologist" of Zhongshan eye center of Sun Yat sen University has entered the community and schools to carry out myopia screening and prediction
in June this year, a big data report on national mobile eye use behavior released by Aier ophthalmology and others showed that as of December last year, the number of Internet users in China had reached 772million, of which 753million were mobile Internet users, accounting for 97.5%. Tens of thousands of netizens who participated in the survey watched the electronic screen for an average of nearly 6 hours a day, accounting for nearly 1/4 of the whole day; they used their mobile phones 108 times a day. That is, the mobile phone will be used every 13 minutes in 24 hours a day. The near impact testing machine for teenagers caused by long-term close eye use is based on the signal generated by the high-speed load measuring sensor at the moment of impact of the impacted sample. After being amplified by the high-speed amplifier, it is quickly converted into a/d digital signal and sent to the computer for data processing. At the same time, it is sent to the computer for data processing by detecting the angular displacement signal, with high accuracy On the basis of the impact structure conforming to the national standard, the instrument is equipped with a high-speed angular displacement monitoring system and a force detection sensor and amplifier. After high-speed sampling and data processing by the computer, it can display N-t and J-T curves, save data, print data reports, etc Visual and other eye health problems are becoming more and more serious
at the same time, in the field of myopia prevention and control, the shortage of ophthalmologists is quite serious. According to the statistics of the ophthalmology branch of the Chinese Medical Association, by the end of 2014, there were about 36000 ophthalmologists in China, less than 4000 optometrists, and 70% of the ophthalmologists in the country were concentrated in large and medium-sized cities. The number of grass-roots ophthalmologists was very small, which was far from meeting the market demand. "Ophthalmologists spend more energy on the treatment of eye diseases and are powerless to prevent and control myopia." Authoritative ophthalmologists in the industry pointed out that myopia prevention and control is more a social project, which is difficult to be borne by doctors alone
based on image recognition technology, ophthalmic artificial intelligence medicine has developed rapidly. Previously, most AI medical products were still concentrated in the field of retinal fundus disease diagnosis. However, AI enterprises are already laying out in advance. Last August, when airdoc, a medical AI enterprise, and the optometry Hospital Affiliated to Wenzhou Medical University launched the optometry AI joint research and development base, the first phase of the cooperation project between the two sides included the intelligent prediction of the progress of juvenile myopia. Zhang Dalei, the founder and CEO of airdoc, told reporters at that time that this project was based on the most complete data set available at present - sufficient and supportable optometry records from the eye optometry Hospital Affiliated to Wenyi University from 2000 to the present, and established an artificial intelligence model to predict the visual acuity progress of year-old myopia children from the current holding sample to the age of 18 according to multiple dimensions. Recently, Zhang Jinglei, vice president of airdoc, told reporters that AI products for myopia prediction have been exploring landing cooperation with a hospital in Beijing
how does AI predict myopia? Professor Lin Haotian, the AI discipline leader and doctoral supervisor of Zhongshan eye center of Sun Yat sen University, pointed out that myopia screening itself is not difficult, and it is difficult to predict myopia when it is adjusted to zero. Everyone has a certain rule in the occurrence of myopia, but how to find out this rule requires high-quality diagnosis and treatment data to learn. Professor Lin Haotian revealed that Zhongshan Eye Center has provided more than 1 million diagnosis and treatment data to machine learning, and a considerable part of the data is continuous. Through in-depth learning and verification, "Ai ophthalmologist" has not only landed in a community hospital in Yuexiu District, but also entered the school to carry out myopia prevention and control work. "Recently, we went to a primary school in Yuexiu District, and the preliminary results of AI prediction show that these primary school students have a relatively high risk of myopia." Professor Lin Haotian pointed out that predicting myopia in advance can advance the occurrence of myopia. It is believed that some smaller factories will be eliminated for intervention
Zhang Jinglei pointed out that the market scale of eye disease diagnosis and treatment in China is about 40billion, while the market scale of eye health such as myopia prevention and control is expected to be more than ten times. Such a huge market potential is expected to attract more capital attention