domeet webmaster

Urgent people need what they need.

Editor’s note: This article is from WeChat public account “Little Leopard Research Center” (ID: cm-ailab), author of the Cheetah User Research Center.

In recent years, with the breakthrough of AI technology such as image recognition and natural language processing, and with the strong support and encouragement of national policies, AI+ medical industry has entered a new era of development. On the other hand, as the problem of population aging continues to intensify, problems such as the difficulty of finding the number one and the uneven distribution of medical resources are becoming more and more urgent. AI+ medical treatment has become a new hope to solve the problem of people’s medical treatment.

This Cheetah User Research Center focuses on the development of AI+ Medical in the fields of electronic medical records, smart guides and virtual assistants, AI medical imaging, etc., and hopes to help readers understand the current impact of AI+ medical industry on people’s lives and current situation. .

PS: The AI ​​medical industry interpretation will be divided into two periods, and the next issue will be the market scan of “AI Health Management”.

Directory

01 Medical Industry Analysis

02 Current AI Medical Layout

03 Electronic medical records of AI Medical

04 AI Medical Wisdom Guidance and Virtual Assistant

05 Medical Imaging of AI Medical

06 AI landing business model

07 Little Leopard Outlook

Part.1Analysis of the medical industry

Medical care is a pillar industry of people’s livelihood, and the industrial chain is extremely long. The medical care in the eyes of ordinary people mainly focuses on the scene of hospitals. The more prominent problems that currently exist in this scenario are the following three points.

• One is from the patient’s point of view, it is difficult to see a doctor, expensive to see a doctor, difficult to register, long waiting in line, no problem after the discharge of the patient.

• The second is that the hospital system cannot effectively allocate resources. The number of hospitals in tertiary hospitals and the workload of medical staff are overloaded, and various resources in primary hospitals are idle and wasted.

• The third is that from the perspective of medical staff, the income is low and the intensity is high. Doctors have a large workload and high risks, and the relationship between doctors and patients is very tight.

At present, the structure of medical services in China is extremely uneven. The annual service number of the 2311 tertiary hospitals reached 1.646 billion, while the number of secondary hospitals was 3.6 times that of the tertiary hospitals.Only “serving 1.178 billion person-times, the number of hospitals is 4.2 times that of the third-level hospitals, only serving 196 million people. The medical services are over-concentrated, the tertiary hospitals are overcrowded, the low-level hospitals are vacant and wasteful, and the grassroots doctors’ income Low, lack of medical knowledge training. The average daily service number of tertiary hospitals is more than 30 times that of lower-level hospitals.

(Source: Statistical Bulletin of China’s Health Care Development in 2018)

Part.2The current AI medical layout

How to improve the efficiency of hospital system operation, improve the accuracy of doctors’ diagnosis, and accelerate the knowledge learning and precipitation of doctors are the hotspots of government and medical industry at all levels. Artificial intelligence technology is the hottest field in the moment. How artificial intelligence technology can provide medical services is the direction that many companies and organizations try. Government management also expects artificial intelligence technology to bring innovation to the medical industry. In the medical field, the upstream and downstream industry chain is long, and artificial intelligence technology has many scenes. Generally, it can be divided into AI-assisted diagnosis, AI health management, AI drug development, and AI disease prediction in a narrow hospital scenario. This article is the first AI medical industry scan conducted by the Cheetah User Research Center, focusing on the application of AI-assisted diagnosis in the medical field.

According to industry statistics, as of the end of 2018, the proportion of artificial intelligence applications deployed in China’s hospitals was only 33.6%, and 42% of hospitals did not attempt to access any artificial intelligence technology. The potential of artificial intelligence technology in the medical field is enormous.

With the development of AI technology and the success of projects in the medical field, the scale of the artificial intelligence market in the medical industry has grown rapidly year by year. In 2017, the market size has reached 13.5 billion yuan, an increase of 40% from the previous month. It is expected to reach 20 billion yuan in 2018. In recent years, the growth rate of the chain has been above 40%.

(Source: Forward Industry Institute)

When we focus on AI assisted diagnosisIt can be subdivided into four sub-fields, including AI medical images, electronic medical records, intelligent medical consultation, and virtual assistants. Patients in the hospital for medical treatment generally follow the following process: registration – waiting – diagnosis – prescription – payment – dispensing, AI-assisted diagnosis can be cut in at all the above links, improve the doctor’s diagnosis speed and quality, reduce hospital operating costs. Most landing solutions now focus on providing solutions to a specific problem.

Part.3Electronic medical records of AI Medical

Electronic medical records are the basis of medical informationization. Electronic medical record information data can be easily exchanged and transmitted across departments, departments, and hospitals. Electronic medical records also provide more convenient sample data for medical big data and disease models. At present, the market size of electronic medical records is also in a period of rapid growth. In 2017, the output value reached 996 million, and it is expected to reach more than 1.2 billion in 2018.

At the same time, the National Health and Health Commission has a clear plan for hospital informationization and electronic medical records. The Health and Health Commission requested that all tertiary hospitals in 2019 should reach the informationization graded evaluation level 3 or above, that is, realize the data exchange between different departments in the hospital. By 2020, all tertiary hospitals must reach grade 4 or higher, that is, to achieve information sharing in the hospital and have medical decision support functions. The secondary hospitals must reach grade 3 or higher.

In the first quarter of 2018, 86.8% of public hospitals in the secondary and above nationwide established standardized electronic medical records (level 1 or above); a total of 8265 medical institutions nationwide registered and participated in data reporting, in the reported medical In the organization, 43.27% of medical institutions can realize data sharing between different departments (level 2 or below) and even different medical institutions.

However, in hospitals that have implemented electronic medical record systems, 60% of hospitals stay in the initial stages of informatization of preliminary data collection and data exchange within the department. It can be seen that the application level of China’s electronic medical record system is still very low, and the hospital informationization market brought by the upgrade of electronic medical records has great potential and the market demand is strong.

Part.4 Intel Medical Smart Guide and Virtual Assistant

The purpose of the Smart Guide is to reduce the pressure on hospital operations and respond to peak patient needs in a timely manner. Relying on the professional medical knowledge map and the local hospital information system, the sensor and human-computer interaction are used to obtain the patient’s body temperature, heart rate and other human parameters, facial expressions, tongue coating and other patient characteristics and disease characteristics, to judge the patient’s condition and complete the consultation recommendation. The accuracy of some current products is about 90%. At the same time, the robot responds in time and saves patient time. Although the guiding robot has been deployed in many hospitals across the country, it is not known whether the efficiency of the operation has been improved. It is mainly limited by the richness and professionalism of the medical knowledge map.

When a patient visits a hospital, the doctor needs to know the basic information of the patient through the historical medical record. This link requires a large amount of time for the doctor, and due to the time relationship, the examination of the historical medical record and the summary of the disease are also lack of systematic. Artificial intelligence technology can precipitate a large number of detailed historical medical records of patients, and can also provide diagnostic help to doctors in combination with disease big data and disease models, thus improving the efficiency of doctors.

Part.5 Medical Imaging of AI Medical

The domestic medical image stock market in 2018 is about 200 billion yuan, and it is mainly in the hospitals above the secondary level. The grassroots level is relatively blank. 200 billion yuan algorithm: In 2018, the total income of domestic medical institutions will reach 4.4 trillion yuan. According to the inspection fee accounted for 10%, of which the ratio of image inspection is 1/3 to 1/2, and the domestic image storage in 2018The market size is 1,500-220 billion yuan. In 2017, the total number of medical treatments in the national medical and health institutions reached 8.18 billion, and the average number of residents visited 5.9 times, which was higher than the number of visits by Americans. However, the conversion rate of image inspection is 20%, compared with the conversion rate of 50% in the United States, there is still much room for improvement. At the same time, with the compression of medicines, the hospital image revenue is expected to continue to grow. In theory, the long-term potential market scale is as high as 30-500 billion. yuan.

At present, the imaging department of the large-scale top three hospitals is in an overloaded state; the medical imaging needs and supply of ordinary tertiary hospitals and secondary hospitals are relatively matched; and the overall service capacity of the primary imaging department is insufficient.

The AI ​​medical image has obvious advantages compared with the traditional artificial recognition image:

A number of viable and highly effective AI techniques have been developed in tuberculosis, fundus lesions, breast cancer, cervical cancer, lungs, heart disease or organs, increasing the likelihood of early detection of diseased lesions. A clearer definition of the disease stage provides doctors with a highly accurate recommendation for a treatment plan.

Part.6Business Model for AI Landing

The AI ​​algorithm for specific scenarios can accomplish tasks efficiently. The main costs of AI medical service providers include production costs and operating costs. Production costs are high, and can be subdivided into computing costs, data costs, and labor costs. The data cost is all kinds of desensitized medical records and medical images, and the medical data that has been meticulously labeled is expensive. Therefore, the AI-assisted diagnosis establishes a close cooperative relationship with data producers such as hospitals to obtain desensitized medical record data for training and iteration of the AI ​​algorithm. When the AI ​​algorithm is stable, the hospital chargeer will be charged a certain usage fee.

Part.7Little Leopard Outlook

Medical care is a civilian production industry, rational allocation, efficient use of limited medical resources, and upgrading of primary hospitalsThe strength of diagnosis and treatment, accelerating the accumulation and learning of doctors’ experience, and responding more quickly to patient appeals and accurate diagnosis have important effects on the operational efficiency of medical systems and the national medical insurance finance. With the development and landing of AI technology, AI algorithm can more accurately identify complex patterns, construct multi-parameter models, and find the characteristics of lesions at an early stage, so as to early detection of diseases and increase the time of treatment.

At the same time, AI combines doctors’ diagnosis or big data operated by hospitals to speed up medical informationization and open up the circulation of data in all aspects of the medical industry. In addition to giants such as BAT, in the AI ​​enabling medical, a large number of artificial intelligence companies such as Keda Xunfei, Yitu Technology, Senyi Intelligent, and Yunzhisheng are constantly exploring the application scenarios and developing AI technologies. Iterative optimization has given the medical industry a new look.