Look at the frontier
Diabetes retinopathy (DR) is one of the most common microvascular complications of diabetes.This lesion is hidden in the initial symptoms, but in severe cases, it may cause permanent vision damage or even blindness. Therefore, early screening and intervention are essential for DR prevention and management.
However, due to the lack of underground filming equipment, scarcity of professional photographers, poor patient screening compliance, DR screening penetration rate is low , and the quality of the film is difficult to guarantee, resulting in poor screening and diagnostic accuracy of related lesions, and it is difficult to effectively achieve disease prevention and control.
In order to solve this problem, Medical Colleges Endocrine Metabolism Department of the Sixth Peoples Hospital of the Medical College, Shanghai Professor Jia Weiping and Li Huating team of key laboratories, Huang Tianyin team, deputy academician of Tsinghua University, Director of the Medical College, Sheng Bin team, a professor of the Department of Computer Department of Shanghai Jiaotong University and the Ministry of Educations Artificial Intelligence Key Laboratory.Deep learning system DeepDR Plus, which is precisely predicting DR progress.
The research results of the cross -cooperative cooperation of medical workers, It is expected to make diabetic patients only take photos in front of a machine to accurately diagnose DRSevere levels can also predict the course of the onset and progress of DR. Related results were published in the international journal "Natural Medicine".
Research team introduced that in the clinical practice of diagnosis and management of chronic diseases such as diabetes, diabetic patients often only screen or follow up in accordance with a relatively fixed time interval, and the complicationsThe exact occurrence or progress time is unknown.This also caused the traditional deep learning model to be accurate modeling of the preface of the progress of the disease, and then it was impossible to predict the individuals onset and disease progress time point. For the first time, the team is based on the vertical line of large -scale medical image. It uses the bottom images and clinical data of more than 200,000 diabetic patients covering more than 200,000 diabetic patients covering multi -national and multi -ethnicians, and innovatively proposes a deep learning framework to successfully realize the risk warning and risk warning of DR progress andTime forecast. The research results can be used to recommend personalized DR screening interval and management strategies, and answer the two key questions faced by clinicians and patients: when will patients refer to the ophthalmology, and the patients DR will haveHow serious.
Study realized the prediction of individualized DR risks and time for the first time.The DeepDR Plus system can accurately predict the individualized risk and time of DR progress in the next 5 years according to the baseline eye image, which is better than the traditional clinical parameter model. The emergence of this research results brings new hope to DR screening and management of diabetic patients, and it is expected to bring more convenient and accurate diagnosis and treatment services to diabetic patients in the future.With the further promotion and application of this technology, it is expected to greatly improve the quality of life of patients with diabetes and reduce the incidence of permanent vision damage and blindness caused by retinal disease.
(Picture Source: Photo Network)
Technical value observation
The upstream of the artificial intelligence industry chain is hardware devices and data devices, including chips, sensors, big data, cloud computing services, etc., providing data services and computing power support for artificial intelligence services Midstream is the technology core of the artificial intelligence industry, including three aspects: general technology, algorithm model, and development platform.> The downstream is an artificial intelligence application product and scenario, which involves multiple fields such as transportation, medical care, security, finance, home, and manufacturing.
The research team uses the vertical queue of large -scale medical imaging, using the eye images and clinical data of more than 200,000 diabetic patients to propose a deep learning framework to successfully realize the risk warning of DR progress.And time forecast. This technology is in the middle of the artificial intelligence industry chain.
Macro market observation
Global artificial intelligence companies are mainly distributed in China and the United States
According to the global corporate database query, the main active enterprises in the artificial intelligence industry are mainly distributed in China. , as of April 2022, China has a total of 6052 active related artificial intelligence enterprises, accounting for 73.9%of the total number of artificial intelligence companies; The second is 609 in the United States , accounting for inquiry7.4%of the total number of artificial intelligence enterprises.
Note: The statistics of China have not included Hong Kong, Macao and Taiwan.
Capital is more inclined to the early investment of artificial intelligence companies
As of October 9, 2022On the day, a total of 6486 investment and financing incidents in Chinas artificial intelligence industry occurred, with a total financing amount of 999.4 billion yuan. Among them, 2014-2018, the scale of financing events and financing scale continued to grow. In 2018, the financing amount reached 136.6 billion yuan and financing events were 1049.From 2019 to 2020, the Chinese artificial intelligence industry market is a lot calm compared to before, and financing incidents have declined but the scale of financing has increased.In 2021, my countrys artificial intelligence capital market once again ushered in a tide of growth, and the number and scale of investment and financing events reached the peak of over the years, increasing to 1066 and 306.2 billion yuan, respectively.As of October 9, 2022, there were 532 investment and financing incidents in the artificial intelligence industry in 2022, and the financing amount reached 100.8 billion yuan.
Note: The data of 2022 As of October 9, the same is the same, and no longer repeated.
Market share: Shangtang Technology Leading Computer Visual Market
According to IDC data, 2022 The size of my countrys computer vision market has reached 12.30 billion yuan.Shangtangs market share has ranked first in the list for five consecutive years. leads the computer vision market, accounting for 23.1%in 2022.Followed by Hainan Vegeous, Innovative Wisdom, Viewing Technology, Yuncong Technology, Smart Eye Technology, and other manufacturers.
The heating of the Chinese artificial intelligence technology track
According to the forward -looking industrial heat map, the current urban clusters related to strong artificial intelligence technologies are mainly concentrated in South China, East China, and Guangdong, Hong Kong and Macao Greater Bay Area, especially Guangdong Province. These urban agglomerations have invested a lot of policies, funds, environment and talent resources for artificial intelligence research and development, and become potential artificial intelligence technology development centers.According to the distribution of thermal maps, the Guangdong -Hong Kong -Macao Greater Bay Area has great possibilities to become a pioneering area for artificial intelligence technology.Focus on related companies in Tianhe District, Guangzhou City, Tianhe District, Haikou City, Hainan Province, and Taijiang District, Fuzhou City, Fujian Province, as well as the investment environment and potential market for the development of the artificial intelligence industry.
Foresight Economist App Information Group
For more industry research and analysis, please refer to forward-lookingThe Industrial Research Institutes "Chinas Artificial Intelligence Industry Development Forecast and Investment Strategic Planning Analysis Report"
At the same time, forward -looking Industrial Research Institute also provides industrial big data, industrial research reports, industrial planning, park planning, park planning, Industry Investment Merchants, Industrial Maps, Smart Merchants Systems, Industry Status Certificate, IPO Consultation/Research Investment Research, IPO Working Draft Consultation and other solutions.In any public information disclosure of the prospectus, the companys annual report and other public information disclosure, the content of this article is required to obtain the regular authorization of the forward -looking Industrial Research Institute.
