North America has become the starting point for Chinese medical AI enterprises to launch their overseas business, and improving the medical system and technology R & D environment will help AI landing

North American countries have rich medical resources, strong scientific research strength, and solid medical information foundation, and have made rapid progress in the application of artificial intelligence to medical treatment. North American IT giants continue to lay out in the direction of medical artificial intelligence by virtue of its informatization strength. Traditional medical device manufacturers and information-based manufacturers have also accelerated the transformation and upgrading with the help of artificial intelligence technology. Hot venture capital further drives the development of medical artificial market. Transnational medical human industrial intelligence cooperation is also accelerating. The medical artificial intelligence industry in North America is ushering in spring. < / P > < p > according to the data of global market insights, in 2018, the global medical and health artificial intelligence market scale was 1.3 billion US dollars. It is estimated that the compound annual growth rate from 2019 to 2025 will be 41.7%, and the scale will exceed 13 billion US dollars in 2025. Among them, the medical and health artificial intelligence market in North America dominated the global market with a price of $653.9 million in 2018. The Canadian government pays attention to the privacy of personal information, the transparency of information and the people’s right to know. In order to ensure the establishment of people’s right to know, and to protect citizens’ personal privacy, PIPEDA stipulates the principles for private enterprises and institutions to handle personal information. < / P > < p > on August 1, 2017, FDA announced the digital health innovation action plan and the software pre certification plan on digital health, and selected several enterprises to participate in the precert pilot. After several improvements, FDA released the pre certification pilot work template version 1.0 in February 2019, as well as the test plan and the latest regulatory framework in 2019. In May 2019, FDA released the recruitment plan to Give more volunteers the opportunity to participate in the test. < p > < p > developed countries have huge medical expenditure and strong willingness to reduce cost and increase efficiency: North American countries mainly include the United States and Canada, both of which are developed countries with high per capita medical expenditure. Therefore, it is an inevitable choice to reduce costs and increase efficiency through artificial intelligence technology. < / P > < p > the foundation of medical informatization is solid, and the resources of standardized database are rich: the information technology industry in North America is developed, the time of medical informatization is early, and there are relatively perfect medical information infrastructure and abundant standardized database resources. < / P > < p > the laws are relatively complete, and the data privacy protection is relatively perfect: North American countries pay attention to the citizen’s privacy, and the laws and regulations on citizen information protection are relatively perfect, and they attach great importance to the privacy protection of medical data, and the data after desensitization is more conducive to research and sharing. < / P > < p > developed medical conditions and strong acceptance of cutting-edge medical technologies: there are many top medical institutions in North America, which are willing to try the latest diagnosis and treatment technology, carry out clinical trials, and become the natural soil for hatching new technologies. < / P > < p > there are many top research institutions, and the industry university research cooperation is close: researchers from Stanford, MIT and other top universities in North America have participated in the research of medical artificial intelligence frontier projects, and have carried out in-depth cooperation with top hospitals and large enterprises, and have achieved remarkable results. < p > < p > venture capital industry is mature and capital strength is strong: developed countries in North America have mature capital institutions and investment chains, and the capital enthusiasm is very active, which can quickly explore excellent medical artificial intelligence projects and carry out capital injection. < / P > < p > NVIDIA Clara platform: Clara is a computing architecture based on NVIDIA Xavier artificial intelligence computing module and NVIDIA Turing architecture GPU. This architecture is supplemented by Clara software development kit, which allows developers to process the data of existing systems by developing various applications and accelerate the R & D process AI consists of 13 classification and segmentation AI, as well as software tools built for radiologists. Clara was launched at the end of November 2018. It is mainly aimed at three types of enterprise customers, the first is medical equipment company, the second is AI software development company, and the third is hospital. So far, hundreds of enterprises and hospitals have registered to use it. In China, Lianying intelligent builds AI architecture on the basis of NVIDIA DGX system, so as to develop artificial intelligence software for full stack medical imaging. It is supposed that NVIDIA Clara platform is also adopted to improve product performance and drive a new generation of AI imaging cluster. < / P > < p > cyclica — drug R & D platform: cyclica’s life design and life express cloud platforms reveal how drugs interact with a variety of protein targets through AI prediction engines matchmaker and poem, helping scientists design, screen and personalize candidate drugs, while improving the safety and effectiveness of drugs. Cyclica’s experiments show that matchmaker, their deep learning engine, achieves greater prediction capability for drug target interactions at the whole proteomics level, with a significantly higher accuracy than molecular docking simulation. Cycla’s machine learning prediction engine poem is used to predict the ADMET characteristics of compounds and to predict the expected behavior of compounds in vivo. < p > < p > Google, University of California, San Francisco – prescription analysis AI system: the system can predict doctors’ prescription decisions in 75% of cases, and can identify patients or prescriptions that appear to be abnormal. The AI system uses a data set for training, which contains about 3 million prescriptions generated from more than 100000 inpatients, using traceable electronic health records with randomly changed dates and deleting some records according to HIPAA. Among them, the data set is not limited to specific diseases or medical fields, which makes the task more challenging, and also helps to ensure that the model can identify a wider range of disease types. The researchers evaluated the long-term and long-term memory recurrent neural network and the logistic model, and compared both models with the baseline. < / P > < p > Yes health – nutrition consultation platform: Yes health combines machine learning technology and computer vision function, not only through text, but also through images to capture, analyze and share data, help users reduce treatment costs and complete personalized treatment. Before eating, patients should take pictures of their own meals, and then upload them to yes health. The platform can identify food types, food consumption and other information from the images, and then convert these information into data that can be used by doctors and patients to monitor the progress of the disease, and automatically generate diet reports for users. Another feature of yes health is to provide automatic messaging services. The platform will use computers to generate message prompts instead of providing manual consultation services. By using AI, which provides health guidance, the platform can provide personalized treatment directly to users’ mobile phones at the right time. < / P > < p > * this is the “core version” of the series of “smart week”. The launch plan of the corresponding “in-depth” version will be announced in the future. Please pay attention to it. For the theme of “Application of artificial intelligence in the medical industry in North America”, what directions or topics do you want to read in the in-depth version of the report? Please leave a message. This will be an important reference for us to make the report in depth. < p > < p > the application report series of AI technology in the heart of machine “smart week” focuses on the traditional industries and scenarios that are undergoing intelligent upgrading with mature application of existing AI technology and great application potential. It presents the industrial landing status, enterprise cases and technology application trends of different AI technologies efficiently and in depth for industrial users by industry and scene by scene 。 The “Chi Zhou” AI application report series contains two versions: the core version and the depth version:

core concentrating essence analysis content to 2 pages, covering important data, analysis conclusions and case descriptions, so as to provide users with efficient and systematic understanding of the potential opportunities of AI technology in its own industry. On the basis of the core version, the in-depth version includes detailed analysis of industry or scene pain points, technical interpretation, detailed explanation of landing cases and data appendix, so as to provide users with in-depth research needs for in-depth exploration. Continue ReadingYueshang group has become the third social e-commerce service platform listed in China after being gathered in pinduoduo