Ai healthcare should be more than just robots and smart image management

     Artificial intelligence medical treatment has long become a popular target for all parties. It is the strategy of the major companies to establish the lead with the technical advantages of the algorithm platform. The reason for targeting the medical field is to see the current situation of numerous hot spots and pain points in people's livelihood: the shortage of medical resources, the prominent contradiction between doctors and patients, and the inequality of medical resources have been the object of ridicule for a long time. From the perspective of artificial intelligence, after the continuous upgrading of technology, finding the landing scene has become the key, and the two sides thus hit it off.

     From the initial emergence of less than 10 startups, now there are hundreds or even thousands of AI medical startups in the country. From the initial no one understood what AI medical does and why, to now the market capital began to actively chase potential stocks, AI medical companies may now embrace their own spring.

     In the field of artificial intelligence, statistics show that among the investment amount of artificial intelligence startups in the medical field, the total investment amount of companies engaged in medical imaging and diagnosis, medical research, medical risk analysis and drug discovery exceeds 80%. In addition, medical robots have become another hot field. This can also be seen from the recent robot conference.

     In terms of image management, it is mainly focus identification and labeling, automatic delineation of target area, adaptive radiotherapy and 3D image reconstruction. From this point of view, it is not strictly the field of AI medicine, or it is mainly targeted by existing standards and medical knowledge. From the perspective of machine learning, it is difficult for computers to independently discover specific images before the onset of disease, or to independently find the rule of early features when breaking through human cognition.

     The wide application of medical robots is more reflected in the role of shopping guide or welcome guests. Even if they provide services through voice interaction function, they still cannot provide services similar to intelligent consultation or virtual doctor. Of course, such services are at a higher stage of artificial intelligence, and they are more difficult. Besides, they face the test of supervision, laws and regulations, and social acceptance. But it also cools the robot, reminding participants to start accumulating user data and digging deeper.

     Still, there should be optimism about the future of both businesses. For example, medical image data is the most common data in medical treatment, and the amount of data is very large. If we can dig deeply into the story behind the image data, it will be of great benefit to the treatment of patients and the future health management.

     The mention of data also touches on the essence of artificial intelligence. After all, only through big data can we train computers to learn independently, so as to find the best solution to break through existing cognition, or provide human beings with services to predict the future and make plans in advance.

     From the perspective of the ideal goal of the future, there is still a certain gap between the current AI players. In view of most market players, most AI medical companies start from a certain part, especially many companies are transformed from Internet medical companies, and most of the medical data collected is uploaded by users actively, relying on previous user groups to support AI data mining. This directly leads to a small amount of data, and most of the single data in a certain field, the content is relatively limited, it is difficult to carry out real depth mining and application service development.

     The driving force of AI medical companies is technology. The key to long-term development is not the traffic in the past Internet era, but the accumulation of data in combination with the characteristics of the AI era, so as to use deep mining to make big data play the function of predicting the future. Depending on the continuous accumulation of data and optimization of algorithm model, for AI medical treatment, after data accumulation, we can make it bigger from the perspective of drug effect analysis and health data management, and comprehensively consider the application services of AI medical treatment after technical support.

     The accumulation of user data is the foundation. From the perspective of strategic layout, it helps AI medical companies to comprehensively target top-level design and consider the overall business layout from the pain points of the medical market.

     To solve the problem of insufficient data volume and single data content, patient data should be obtained from multiple sources. Besides public hospitals, physical examination centers and research institutions can also cooperate to obtain patient data. Secondly, it is necessary to open up market-oriented channels with insurance companies, research institutions and enterprises, apply the results of artificial intelligence mining to the general public, and contact the C end with the help of the B end. Finally, it is still necessary to continuously improve the level of technical support, optimize algorithms and models, and let AI provide more future prediction services.