1. Diversified medical artificial intelligence products and strong downstream demand

The upstream of the medical artificial intelligence industry is mainly industries that provide basic technical support for the industry, such as medical data mining, algorithms, etc. Typical companies include Carbon Cloud Intelligence, Lianxin Medical, and Dashu Yida.

The downstream is mainly the application layer of medical artificial intelligence technology. The main application scenarios include medical imaging, virtual assistants, drug research and development, health management, disease risk prediction, medical history/document analysis, and representative companies include iFLYTEK, BGI, Tumar Shenwei, Boao Biology, etc.

In the medical artificial intelligence technology industry, a large number of enterprises have emerged in my country. These enterprises have developed various medical artificial intelligence products for the medical industry, such as intelligent imaging-assisted diagnosis and treatment systems, diagnosis and treatment robots, and voice electronic medical records.

Chart 1: Medical artificial intelligence industry chain map

In terms of downstream demand, China’s medical industry has long suffered from uneven distribution of high-quality doctor resources, education reforms on the rate of misdiagnosis and missed diagnosis, high medical costs, and large gaps in the supply and demand of doctor resources. This problem has become more prominent as China's medical reform is gradually advancing and hierarchical diagnosis and treatment is gradually implemented.

In addition, the distribution of medical resources in my country is severely uneven. The top three hospitals are mainly concentrated in Beijing, Shanghai, Guangzhou, Shenzhen and provincial capital cities, and a large number of high-quality medical resources have been concentrated, causing a large number of patients to flood into these cities. Most of the time, large hospitals treat a large number of patients with common and frequently-occurring diseases, which makes it difficult and expensive to see a doctor, wastes valuable medical resources, and makes it difficult for severe and emergency patients to receive timely treatment.

In this context, artificial intelligence will increase the level of medical services and improve the current status quo under the combined effect of various scenarios. According to statistics, less than 50% of Chinese hospitals have deployed artificial intelligence applications and maturely used them, and nearly 50% of hospitals still have not tried to access artificial intelligence technology.

2. AI+ medical policies boost the rapid development of the industry

In recent years, the state has issued a number of national policies and special policies for medical artificial intelligence, which fully reflects the importance attached to the AI+ medical field at the national level. In April 2018, the General Office of the State Council issued the "Opinions on Promoting the Development of "Internet + Medical Health"," researched and drafted by the National Health Commission and other ministries. The "Opinions" released multiple policy benefits: 1) Recognizing the legitimacy of "Internet hospitals"; 2) Supporting the outflow of prescriptions; 3) Promoting "Internet +" medical insurance settlement services.

In addition, the research work on "confirmation" of medical electronic information is underway; the "Opinions" clarify that by 2020, hospitals above the second level will generally provide time-based appointments for diagnosis and treatment, intelligent medical guidance and triage, waiting reminders, inspection and examination results inquiries, and diagnosis. Online services such as time settlement and mobile payment.

Driven by national policies and the needs of various medical fields, the market size of my country's medical artificial intelligence industry has been rapidly expanded. The compound growth rate of 2017-2019 reached 31.98%, which shows that the industry has great potential. According to statistics from the Chinese Institute of Electronics, the market size of the medical artificial intelligence industry will reach 7.53 billion yuan by 2021.

Chart 3: 2015-2018 China's medical artificial intelligence industry market size change trend (unit: 100 million yuan, %)

3. The core application value of each end of medical artificial intelligence is clear

Medical artificial intelligence uses a wide range of application scenarios in the patient, billion yuan, and ecological segments to analyze the core application value of medical artificial intelligence from different angles. Specifically:

①Patient side: Artificial intelligence reshapes the medical experience. Using artificial intelligence technology, medical services can break through the physical boundaries of the hospital, center on the patient, and extend to the whole process of medical treatment before, during and after the diagnosis.

②Hospital end: Artificial intelligence reconstruction management system. Artificial intelligence in-depth patient management (electronic medical records), medical equipment management (intelligent closed-loop management of equipment and drugs), ward management (intelligent surgery scheduling), performance management (DRGs performance), backstage Management (intelligent back-end comprehensive management such as human, financial and taxation) and other aspects have brought the overall upgrade and reconstruction of the hospital management system.

③Ecological end: Artificial intelligence improves the medical service ecology. In the entire medical service system, hospitals are at the core and are the intermediate hubs for the aggregation and integration of various information and data. In addition, there are other medical service institutions, medical and health product providers, and payers. , Regulators, etc.

Figure 4: Analysis of the core application value of artificial intelligence in the field of hospital management

4. The product layout of domestic medical artificial intelligence companies is showing a decentralized trend, mainly concentrated in two major areas

It has been 7 years since Watson established its medical business development direction in 2011. In the past 7 years, artificial intelligence has developed rapidly around the world. Countless companies have followed the trend, and deep learning algorithms have undergone many changes. But after the wave, a large number of companies also closed down.

Now, companies that have survived on a global scale have gradually formed a head camp in the field of medical artificial intelligence. Under the deep learning process, each company can report an accurate and high number for their AI products. However, the new era is no longer a one. In the era of algorithms alone, the quality of AI is no longer judged by a number or the result of a human-computer war. Therefore, only if artificial intelligence technology is widely used in hospitals can it survive for a long time.

At present, the domestic AI medical industry's competition is focused on the landing link. From the perspective of active medical artificial intelligence companies in the market, the products are mainly deployed in 10 major fields such as medical imaging, medical record/document analysis, health management, hospital management, and virtual assistants.

Chart 5: Distribution of product lines of domestic medical artificial intelligence companies

From the product research directions of these 108 companies, it can be seen that lung nodule screening and sugar network screening are the two most popular directions. There are 33 medical artificial intelligence companies assembled, accounting for about 30.6%; but there are also quite a few. Many companies have set their sights on cardiovascular diseases, and the number of companies has exceeded 10. It can be seen that the products of domestic medical thermal intelligent companies are showing a decentralized trend.

Chart 6: The distribution of the number of domestic 108 medical artificial intelligence companies engaged in product research and development for various diseases (unit: home)

From the perspective of the specific layout of various medical artificial intelligence products, the two popular products, medical imaging and disease risk prediction, are more concentrated. According to statistics, there are currently 43 companies providing medical imaging services, mainly Alibaba Cloud and Wingspan. Technology, Xinjian Medical, etc.; 45 companies provide disease risk prediction services, such as Tuma Shenwei, Berry Health, Boao Bio, etc. The following table shows typical domestic medical artificial intelligence products and major layout companies.

Figure 7: Typical domestic medical artificial intelligence products and major layout companies

5. Medical robots are widely used, and rehabilitation is the main field

Medical robots are one of the subdivisions of robot applications, specifically referring to medical or paramedical robots used in medical scenarios such as hospitals, clinics, rehabilitation centers, etc. According to the International Federation of Robotics (FR) classification, medical robots are further divided into surgery Robots, rehabilitation robots, auxiliary robots, and medical service robots are four major application areas.

Chart 8: Analysis of main brands and application characteristics of medical robots

According to CCID data, in my country's medical robot market in 2019, rehabilitation robots, surgical robots, assistive robots, and medical service robots accounted for 47%, 17%, 23%, and 13%, respectively.

Chart 9: China's medical robot application field structure in 2019 (unit: %)

6. China's medical artificial intelligence industry is favored by capital, but there is a phenomenon of agglomeration of investment

From the perspective of investment cases, in the first half of 2012-2020, the popularity of equity investment in my country's artificial intelligence medical field showed a trend of first increasing and then decreasing. Among them, the years of higher equity investment were 2016, 2017 and 2018. In 2019, affected by the cold winter of capital, the investment in artificial intelligence medical field has dropped, the number of investment cases has fallen to 52, and the investment amount is 1.546 billion yuan.

In the first half of 2020, due to the impact of the new crown epidemic, the number of investment cases remained at a low level, and there was no obvious rebound. The number of investments in the first half of the year was 14 and the investment amount was 917 million yuan.

Figure 10: 2012-2020 equity investment scale and number of cases in the artificial intelligence medical field (unit: 100 million yuan)

According to incomplete statistics from the health industry, from January to December 2020, a total of 65 financing incidents occurred in the domestic medical and health AI field, of which medical imaging accounted for about 1/3 of the total financing. From the perspective of financing rounds, some companies have begun to enter the mature stage of development, and companies whose product technology has been recognized by the market and their business models have become more mature are also more likely to be favored by capital. Various enterprises have also increased their technological research and development efforts to build competitive advantages with innovation and accelerate the pace of development.

Figure 11: A list of my country's AI+ medical financing events from 2020 to 2021

In addition, in 2020, the passing of the three types of artificial intelligence devices is a milestone event for the medical AI industry. In 2020, a total of 9 AI medical imaging products have passed NMPA approval. The AI medical imaging industry has passed the review and approval stage that has troubled it for many years, and has moved from "application landing" to "commercialization".

Chart 12: 2020 "AI+Medical" NMPA approved products

7. The epidemic will accelerate the development of medical artificial intelligence

The application of AI in medical treatment is becoming more and more extensive, especially the medical imaging-assisted diagnosis system, which has covered the screening of a variety of common cancer types, and the product itself can already be embedded in the doctor's workflow. Under the catalysis of the epidemic, AI + public health, AI + hospital management, AI + medical robots, AI + drug research and development, AI + health management and AI + medical payment will be further developed.