How will AI+ medical images be shuffled after the dialogue?
On June 15, 2018, under the guidance of Shanghai Economic and Information Technology Commission, Shanghai Municipal Commission of Commerce, Shanghai Changning District People's Government, Shanghai Changning District Youth Federation and Yiou Company jointly organized "2018 Global Intelligence + New Business The Summit - Smart + Great Health Summit was held at the Shanghai Changning World Trade Center. This summit will focus on AI and medical care , and will conduct a full and in-depth discussion on AI-enhanced medical care around digital life, smart medical, genetic testing, AI imaging, health management, and hospital management.
Guests attending the summit included Ma Jun, Dean of Shanghai Tongren Hospital, Kang Rong, Vice President of Microsoft Greater China, Wang Xi, Vice President and Chief Technology Officer of Philips China, and Founder and CEO of Tumar Shenwei, Founder of Voxel Technology Ding Xiaowei, CEO and CEO, Zhang Chunxi, Vice President of Technology Marketing, Li Chaoyang, Vice President of Shenrui Medical Market, Zhao Nan, Co-founder of Jellyfish Gene, and Li Yuxin, Founder of Health and Benefits, Sun Qi, Founding Managing Partner of Dow Investment, and Deputy Director of Yiou Company President Gao, Vice President of Yiou and Dean of Yiou Think Tank Institute by Tian Yu.
At the "Smart + Great Health" Summit, Ma Jun, Dean of Tongren Hospital, Shanghai Jiao Tong University School of Medicine, Zhang Chunyu, Vice President of Technology Marketing, Sun Qi, Founding Managing Partner of Dow Investment, and Vice President of Yiou and Research Institute of Yiou Dean started a roundtable discussion on the topic of AI+ medical imaging by Tianyu.
The following is a live discussion shorthand:
By Tianyu: Hello everyone! I am the head of the research business of the company, and this roundtable discussion has been waiting for a long time. We invited Ma Jun, Zhang Chunxi and Sun Qi as guests at the roundtable discussion.
Why are you looking forward to this round table? I believe that everyone can see that the three parties sitting on the stage today represent three important participants in this field, namely representatives of the hospital, representatives of medical technology companies, and representatives of investors in the medical field. Please ask each of the three to make a simple self-introduction.
Ma Jun: Good morning everyone! I am from Shanghai Tongren Hospital, also known as Tongren Hospital of Shanghai Jiaotong University School of Medicine. I am Dean Ma Jun. The first feeling I was sitting here today is that I am too old. All the guests on the stage today are young and young, and all the audiences are very young, so I am too old. I have been working in the medical industry for nearly 30 years and have worked in management positions for 20 years. Medical informatization has gone through 20 years of history. The hot Internet + medical model and the decision support based on big data and the artificial intelligence that last year have been around the past few years. Today's topic is about medical imaging. The medical image-related artificial intelligence team is communicating and cooperating with us. I am also looking forward to this exchange and sharing. I hope that more ideas and ideas will be brought back, thank you!
Zhang Chunxi: Hello everyone! I am Zhang Chunxi of Beijing Pushing Technology Co., Ltd. Beijing Sixiang Technology is a leader and pioneer in medical imaging and artificial intelligence. Beijing Pushing Technology has been in existence for three years since its establishment in March 2015. We now have many users in the top three hospitals in the country. At present, Imagine Technology is headquartered in Beijing, and now has branches in Japan, the United States, and Germany.
Sun Qi: Hello everyone! I am Sun Qi, managing partner of Dow Investment. Dow Investment is an early investment institution focused on medical health. We are also a company in the medical professional GP that has less continuous layout in medical AI because most medical care GPs are more cautious, and investing in artificial intelligence is more of a TMT or a comprehensive investor.
In 2016, he began to invest in Wuhan Landing. Landing is the largest in China in terms of pathological diagnosis AI, and it is also rare in the medical AI to achieve large-scale commercialization and revenues of over 100 million. Landing uses artificial intelligence to do Cervical cancer screening, remote color reading diagnostics. He also invested in Shenrui, a platform company in the field of medical imaging, and is now firmly in the first echelon. More than 300 million financings were completed in just one year. Later, he voted for a brain doctor. The brain doctor made artificial intelligence diagnosis in Alzheimer's disease. Haier Capital also increased its holdings in the past few months. Haier is focusing on Kangyang. Therefore, I have had the privilege of making some attempts in the field of medical AI, and I will share more with you later.
By Tianyu: You can see that the three parties you are visiting today are very representative. Tongren Hospital is a hospital located in Changning District. It is one of the companies in the head of artificial intelligence medical imaging in China. The investment in medical care is one of the few companies, and the investment is also very successful.
The first question is first thrown to Zhang Zong, who thinks about technology. How many years have you seen the impact of artificial intelligence on medical imaging?
Zhang Chunxi: There are a lot of seniors and colleagues in AI. From the perspective of artificial intelligence, there are two aspects in medical application: the first breakthrough in the algorithm of deep network learning; the second aspect is the application of deep network learning in medical imaging. In the past, there were some so-called artificial intelligence called traditional CAB. This technology is a so-called expert feature learning technology, which is characterized by slow speed and low accuracy. Deep learning can bring more accurate computing time and high efficiency.
By Tianyu: Let this topic continue to the Mayuan side. You feel that the current AI deep learning technology is different from the past AI in clinical applications. What are the changes?
Ma Jun: The constant is the diagnosis. We all expect that with the continuous improvement of AI, we can improve the sensitivity and accuracy. Everyone is working hard in this direction. What has changed is that the speed and application fields are broader, and some areas that were unimaginable are infiltrating. In addition, it is not only in the field of diagnosis, but also in the field of diagnostic decision support and drug prescription. So this is also very worthy of our expectations.
By Tianyu: What are the cases in Tongren Hospital that are already in practice or cooperation?
Ma Jun: It is still quite extensive now. In particular, I sat down here and listened to the reports of the previous guests. In fact, we have had shallow or deep contact with many teams in this field, such as In terms of the diagnosis and decision support system, our knowledge base and self-service robot triage were introduced immediately. In the field of imaging, at least two companies are talking to us now, and some databases have been imported for automatic triage and picking. In addition, in the pathological diagnosis, there has been communication with some teams, and there is gene sequencing . Recently, a company is talking to us about the application of chronic disease management in the field of artificial intelligence.
These jobs are just getting started, but in fact there are still a lot of clinical applications.
By Tianyu: When you cooperate with external companies, what are the requirements and conditions for them?
Ma Jun: The first article of the dean is that you can't talk about money for free. You have to rely on us to grow up. So the first thing is to talk about the economy. We can't let us contribute resources and data. The second is data security. This is a very difficult thing. It is very entangled in management's decision-making. What about ethical and patient-based privacy protection? What about knowledge patent sharing? The third is that the clinician is both open and accepting and feels a little troubled, because the ideal is to improve efficiency, but there are not many software applications that can really improve the performance. We and the experts in the process of cooperation And clinicians are contributing their wisdom and how much they affect the efficiency of their work. Therefore, in the process, we must ensure that our daily work is not affected and the project is carried out in an orderly manner. The fourth article is the policy. Today, the dean was invited to come over. Without the secretary coming over, the health authorities are still cautious about this field, so they must be reasonable and legal.
By Tianyu: This question was thrown back to Mr. Zhang. What are the requirements and conditions for Ma’s Dean, how do you deal with cooperation with different hospitals? Is there any difficulty?
Zhang Chunxi: First of all, our products are used in hospitals as new products. The biggest contradiction encountered now is a new one. As before, I used to write by hand. Now suddenly I have become a lot of people typing on the keyboard. I believe many colleagues. Have a deep understanding. This is a very difficult and long process, but it has been widely used in more than 300 hospitals across the country, not only in physical examinations, but also in outpatient and inpatient settings. Many hospitals have developed a strong sticky habit, especially when the computer network is not good. Many doctors in some hospitals feel very uncomfortable after using our AI. They have been asking us why we are still not doing well. It is very good.
In the same situation, there are many small hospitals that are very interested in our products. Many of us are talking about marketing cooperation. In terms of national policies, the CFDA has not yet approved it.
By Tianyu: Dow Investment as an investor, how does Sun always think about the rhythm of AI medical imaging in the next few years? And how do you judge whether a company can actually land in a hospital and do better?
Sun Qi: I think a few key nodes. Just Ma Dean also said that the use of AI is still very inconvenient. This is very objective and true. It has been more than two years since the emergence of medical artificial intelligence startups. In the past two years, we have been working with some head hospitals to establish a database of scientific research data and to polish the products step by step. At the end of last year, Shenrui was the first to receive a second-class license. Now some manufacturers have obtained the second-class license. The third-class license has not been available so far, because the CFDA standard database has not been built yet, and it will be the next year at the end of this year. Can be approved at the beginning of the year.
Now that the capital is bustling, the government is also very excited to be on the surface, but the bottom line is to speed up product research and development, newspapers and the occupation of the head hospital. Now that you haven’t really competed, what are the next nodes that take the CFDA three-category certificate? How many families can get it? Some can't get it when they can't. Further down is the question of the charge list. What form do you charge after you get the certificate? Now the AI ​​that is invested is to do software and hardware integration, like Shenlan is not medical AI, they are retail, they did not get the certificate. So in addition to technical solutions, they can also be combined with unmanned retail machines for hardware and software integration.
However, medical thresholds are higher and there are government policies. If you get the certificate, you can still charge the government, unless you go around doing software and hardware integration. Then look at everyone's marketing ability, I think there will be no more head willingness, because the head will be very cautious to use artificial intelligence, the initial AI products will still be in the lower stage. It is still in the enthusiasm stage, and the big competition has not yet arrived.
By Tianyu: When did you get the second-class certificate?
Zhang Chunxi: Imagine the first two types of certificates.
By Tianyu: Almost all of the time nodes get the second class certificate.
Zhang Chunxi: The second type of certificates may be batches, not one.
By Tianyu: I think that during the cooperation with the hospital, can Zhang always tell us what is the biggest difficulty you have encountered?
Zhang Chunxuan: There are two difficulties. The first is the doctor's habits. Secondly, AI is now a hot spot. The AI ​​is flying all over the place. Everyone is really good. It is like a customer. The customer says that the customer says you. Good is good. At present, it is assumed that there are already 100 top-level top three hospitals in the country, and many hospitals are highly dependent on the assumptions.
By Tianyu: Mayuan, if there are 5 or 6 companies that do the same service at the same time, how do you compare or PK out which one is better? Which is my preferred partner?
Ma Jun: Don't worry, I actually want to jump out of this question to answer. I think that doing artificial intelligence and doing medical care is probably the most intellectually and mentally developed group in the society. Sitting together can surely collide with the spark of wisdom, specific business negotiations. The problems with the product will be easily solved, and I am personally confident about the future development of this field.
In addition, I would like to start a little. The boss who just thought about it should not underestimate the acceptance of modern technology by medical staff. Doctors are the easiest to accept the progress of modern science and technology. Every progress we try is to embrace and open mind. It is precisely because of this mentality that when the AI ​​is not so mature in the initial stage, more than 100 tertiary hospitals have begun to cooperate with you, which means that everyone is willing to support your growth.
Doctors have a high level of wisdom in new technologies and medical integration. Many experts have more comprehensive intelligence than management. They will definitely provide very professional and intelligent advice for AI applications in the AI ​​field, so I think As long as our two fields are open and well applied and communicated in practice, the future development of this field is very promising.
There are three concerns from my personal point of view:
1. Because the disease model has changed from the past “biological model†to “biological psychosocial modelâ€, it has turned from disease-centered to human-centered. So today's "health" is a false proposition. If you have been around the disease, what about the person's problem? What about the management of comprehensive diseases and integrated diseases? This is what the medical industry has to think about.
2. The growth of doctors needs to be accumulated in practice for generations. Yesterday, when I was in communication with the directors and the directors, they were very worried that the younger generation of doctors would abolish. He would not have the basic training in the early stage, that is, the fool-like treatment, once complicated When the situation returns to the original logical thinking stage, there is no training. I think a company like Philips should think about artificial intelligence training.
3. In all the reports today, the future AI will improve efficiency and reduce costs, which is what the hospital director needs to accept. However, from a macro level, the cost of medical society is getting higher and higher, because it is easier to see a doctor, because it is easy to see a doctor, so over-medical care will accompany it. How to control social health from a macro perspective is something government departments need to think about. So they have restrictions on us in the policy area.
By Tian Yu: Ma’s thinking about the problem is very high, jumping out of the scope of simple technology impact, because technology will bring progress and other changes to the mechanism.
Then pull the problem back to explore a detailed issue related to artificial intelligence, data resources, data acquisition problems. I believe that all parties are very concerned about this issue. Today, the leadership of the Changning District Health Planning Commission is also in the government's coordination and management of data resources. Participants are invited to talk about the data of artificial intelligence in the process of medical image landing. Is the acquisition of resources a problem? Is there any good way to crack it?
Sun Qi: We believe that medical artificial intelligence is very important in the early stage of this stage. The most important ones are products and scenarios. One is the ability of data technology at one end. However, with the promotion of the government and social market education, continuous access to high-quality data in the long run is what all medical AI companies want to do, but these problems are slowly improving, and in the long run, it is not a big problem. Including the country is building three national teams for medical data, and integrating the scattered data of each hospital by the national team. He completed the "highway" and then repaired the "small national road" application scenario. The government has already done something. Entrepreneurial companies think that technology will not wait for the government to repair the road to do it again, and the opportunity will be gone. Before that, everyone is doing cooperation with the head hospital and the top three hospitals to obtain desensitization data in these ways. Hospital cooperation, this is a very feasible method at the current stage.
By Tianyu: What is the experience of the technology of thinking as a practitioner? Is there any solution to the problem?
Zhang Chunxi: Imagine that science and technology strictly abide by the data network law of the People's Republic of China, and our servers are constantly optimizing and iterating in hospitals. This is one aspect. On the other hand, through deep research cooperation with many top A hospitals, and even deep cooperation to do some model development, now the national policy is also encouraging the integration of production, education and research, we are starting from these aspects.
By Tianyu: Does Ma Dean have a better solution at the management level of hospital data?
Ma Jun: From the government level, we should give us more clear rules. Now we are all making rules. The hospital has also done a lot of certification. The government should give us rules, what data is used in what form, what is allowed and which is not. Allowed, I just said that this is an industry requirement. The government has to invest in us, so that we can pay back in the process of sharing and sharing data, so that we can use this data more quickly. This is what we ask the government.
By Tian Yu: Dean Ma also raised some expectations for the leadership of the Health Planning Commission.
The final part of the question discusses the judgments and opinions about the future. Dow Investment Sun, you invest in a large number of companies in the medical field, some are pan-medical, and some are related to AI, how do you judge the development opportunities of artificial intelligence medical? In addition, what are the possibilities of other tracks?
Sun Qi: From the perspective of medical artificial intelligence, the first is patience. The introduction of the supporting measures just mentioned by Ma’s director must be patient, because the government has attached great importance to medical artificial intelligence. In addition, we must have patience in the gradual improvement of products. The accumulation of time is 2 or 3 years. In 3 or 4 years, the products are more and more successful. When we can provide a package of technical solutions, I believe that the doctor's experience will be much better, so be patient. For investors, for the medical AI companies, from the good vision of everyone to the final large-scale landing, commercial realization, profit, we must also have patience, so patience is very important, this is the right thing is a broad At the track, we have to accompany the invested companies to grow and wait for the development and prosperity of the industry.
The medical industry has invested more in the field of surgical robots. It feels that the future of large surgical instruments is based on surgical robots. Minimally invasive surgery is the mainstream of future surgery. Of course, surgical robots also have points that can be combined with AI. . In terms of medical services, the focus is also on investment layout, and there is a lot of attention. Most startups and investors pay attention to technology, technology and institutional change are the two main lines of entrepreneurship and investment. The transformation of technology is happening day and night, so everyone is paying attention. The reform of the system is very important. The main line, but not every time and anywhere, has the opportunity. If there is no new round of medical reform and consumption upgrade, the time window of institutional reform will not appear. From now on, the next three to five years will be promoted by institutional reform, new medical reform and consumption upgrading. The medical service that has brought prosperity has developed a very good time window, which is our focus and layout.
By Tianyu: Tell us about your time node from the perspective of the future of the technology. You think that the years are a step, and in a few years, it is the next step to achieve large-scale landing and commercial income.
Zhang Chunxi: From the point of view of the imagination, it has been very perfect, and investors often have meetings and pressures.
By Tianyu: Roughly speaking, your prediction of the time node.
Zhang Chunxi: First of all, I will definitely wait for the national policy CFDA to come down. From the perspective of imagination, there will be a large area of ​​sales. There is no doubt about this. If you sell well, then it is another matter.
Sun Qi: I have just mentioned three nodes in the three mountains, 1. Approval of the government CFDA three types of certificates. 2. Charge; 3. When "a horse is flat", see who is running fast, and fight for operational ability.
By Tianyu: Dean Ma, do you think that a few years is what you expect or may happen to the hospital and large-scale application?
Ma Jun: In the past, I have been in the IT history for 5 years, but in the current era, I want to think for 3 years, but now I feel that it can be 2 years. We must learn to wait, but we don't want to wait too long, so that professionals can be freed from the heavy, trivial, repetitive, and inefficient work conditions as soon as possible.
By Tianyu: Two years later, if you organize another discussion, I don’t know if it has achieved relatively rapid development and commercial landing at that time, let us look forward to it!
Today’s round table is over, thank you for listening, and thank you for the participation of the three guests.
thank you all!
Diabetes is a group of metabolic diseases characterized by high blood sugar. Hyperglycemia is due to insulin secretion defects or their biological effects of damage, or both caused. Hyperglycemia in long-term diabetes, leading to various tissues, especially the eyes, kidneys, heart, blood vessels, nerve chronic damage, dysfunction. The etiology of diabetes is roughly one is Genetic factors. Type 1 or type 2 diabetes had significant genetic heterogeneity. There is a family history of diabetes mellitus, 1/4 ~ 1/2 patients with family history of diabetes. Clinically at least 60 kinds of genetic syndromes may be associated with diabetes. Second, Environmental factors, eating too much, reduce physical activity caused by obesity is the most important environmental factors of type 2 diabetes, so that type 2 diabetes genetic susceptibility to individuals prone to disease. There is no cure for diabetes, but through a variety of treatment can control diabetes. Mainly includes five aspects: diabetes education, self-monitoring of blood sugar, diet therapy, exercise therapy and drug treatment. Antidiabetic drugs are needed in the patient after eating and exercise therapy and diabetes health education, blood sugar control can not be achieved when the target treatment. Most antidiabetic drugs have a greater side effect, so the patient must follow the doctor's instructions. With the succession of new antidiabetic drug patents,the patient is not only actually reducing the cost of treatment, but also increased the number of treatment options, Take greater hopes to patients with disease control.
Anti-Diabetes Intermediate,Diabetes Mellitus,Diabetic Drugs,Newer Antidiabetic Drugs
Taizhou Volsen Chemical Co., Ltd. , https://www.volsenchem.com