According to the annual report on global AI investment trends released by CB insights, global AI start-ups raised $26.6 billion in 2019, while healthcare accounted for us $4 billion. Healthcare is still one of the hottest areas for AI landing.
On the other hand, in 2020, the pace of “new infrastructure” construction will start to accelerate.
Prior to this, Beijing, Fujian, Henan, Yunnan, Guangdong, Jiangsu and other places have issued a list of major project investment plans for 2020. According to incomplete statistics, the total scale of investment plans for “new infrastructure” has exceeded 40 trillion yuan.
Medical treatment is closely related to people’s life and health, and “medical + new infrastructure” naturally receives special attention. Artificial intelligence is an important part of the new infrastructure construction. Leading enterprises in the medical industry, such as Keya medical, inference technology, Huiyi Huiying, voxel technology, naturally become the “backbone” of this construction wave.
A common feature of these enterprises is that they were established almost at the same time (2016), which was called the first year of artificial intelligence. In more than four years, they have proved their technical and commercial strength to the industry.
At the end of the 19th century and the beginning of the 20th century, Italian economist Pareto put forward the “28 th law”, also known as the key minority rule, which is also applicable to the medical AI industry.
The future thinking of medical AI head enterprises will greatly affect the trend of the industry.
Keya medical: using three kinds of certificates to correct the name of the industry
As the first enterprise in the medical AI industry to obtain the medical AI class III certificate, Keya medical in January this year really boosted the confidence of the medical AI industry.
Keya medical said that the new medical infrastructure is a process of intelligent medical construction based on medical big data platform and a new generation of medical information network, with artificial intelligence technology as the core driving force.
Under the influence of “new infrastructure”, the future medical AI industry will pay more attention to the following three aspects:
1、 Clinical value of products: the core purpose of medical AI technology is to serve the clinic, so the clinical value of products in the whole clinical process application will become an important principle to test the product value.
2、 Effectiveness of results: Medical AI is actually the intelligent collection, management and analysis of multi-dimensional and full cycle data, so as to realize multiple functions such as discovery, early warning and auxiliary decision-making, and the result effectiveness feedback of these functions will become the only standard for verifying products.
3、 Product compliance: the launch of the new infrastructure has given the medical AI industry a lot of opportunities. At the same time, it is necessary to introduce a supporting management mechanism. Therefore, as a standard test and threshold for the industry, product compliance will become particularly important.
Lei Feng learned that at present, the first product of Keya medical is “coronary flow reserve score calculation software (deep pulse fraction +)”.
“Deep pulse score” is the first class III artificial intelligence medical device registered and approved by the State Drug Administration (nmpa) in China. It is also the only product with dual certification of China nmpa and EU CE. The State Food and drug administration made a qualitative evaluation on it with “significant economic and social benefits” in the announcement. At present, there is no product of the same kind registered on the market in China.
Deep pulse fraction adopts the deep learning technology created by Keya medical, which intelligently optimizes all links from medical image processing, model reconstruction to FFR calculation, effectively reduces the influence of image noise and artifacts, improves the accuracy and accuracy of segmentation, and greatly improves the processing speed.
Compared with the current screening method based on observing the stenosis degree of CTA (lumen diameter stenosis rate ≥ 50%), the diagnostic accuracy, specificity (FFR 88.6% vs CTA 30% – 40%), positive predictive value and negative predictive value of deep pulse fraction have been greatly improved, which can effectively reduce the false positive probability.
Keya medical said that the clinical application of deep pulse fraction will benefit patients, hospitals and society
For patients, DPF is noninvasive, accurate, convenient and efficient, which is conducive to the early diagnosis and multiple reexamination of patients, avoiding over treatment (unnecessary invasive coronary angiography and subsequent PCI), and relieving the pain and economic pressure of patients;
For doctors, deep pulse score is embedded into doctor’s existing workflow to take over the tedious, repetitive and low monomer value data screening behavior, so as to help doctors quickly and accurately extract valuable information from clinical data and obtain real and accurate quantitative evaluation indicators, so as to help formulate individualized treatment plan and promote the speed and efficiency of medical process;
For the society, deep pulse score can significantly save medical resources and medical expenses, reduce social medical costs, and has significant economic and social benefits.
At present, Keya has cooperated with more than 200 large tertiary hospitals, including Beijing Anzhen Hospital Affiliated to Capital Medical University, general air force hospital of Chinese people’s Liberation Army, Qilu Hospital of Shandong University, Southwest Hospital of the Third Military Medical University and Guangdong Provincial People’s hospital to jointly build the “deep pulse fraction artificial intelligence diagnosis center”, and develop a series of vascular function assessment around the circulatory system Products and intelligent imaging platform for imaging department.
Focusing on the “new medical infrastructure”, in the coming year, Keya medical will focus on the clinical application of artificial intelligence technology in clinical medicine, accelerate the application of AI medical products from medical technology departments to various clinical departments, integrate multi group information including medical records, imaging, pathology, biochemistry, immunology, molecular diagnosis, etc., and serve for early screening, diagnosis and treatment Clinical application scenario, truly realize the whole process coverage of single disease.
Speculate on science and technology: two “representative works” devoted to new medical infrastructure
Chen Kuan, founder and CEO of science and technology, believes that the new medical infrastructure will have a huge impact on the medical industry, especially the construction of medical infrastructure. It will strengthen the infrastructure construction in clinical, hierarchical diagnosis and treatment, public health and other fields, so as to have a more efficient and powerful response ability.
Taking Xinguan epidemic as an example, the emergence of new technologies such as artificial intelligence has greatly enhanced the ability of monitoring, case screening and diagnosis and treatment of Xinguan epidemic, which has provided great help for rapid response and control of the epidemic situation.
At present, it is assumed that there are medical products and solutions in many fields, including two “representative works”.
The first is the inference technology pneumonia intelligent auxiliary screening and epidemic situation monitoring system.
The novel coronavirus pneumonia is coming to the forefront. We hope that technology will cooperate with several top hospitals in Wuhan, and we will iterate and optimize the special image performance of the new crown pneumonia. Novel coronavirus pneumonia can help doctors detect lesions and alert early, and avoid rapid diagnosis and reduce risk of cross infection in new cases.
On the novel coronavirus pneumonia, the patients can be diagnosed and diagnosed by the intelligent quantitative analysis of the abnormal cases and the severe cases. The monitoring and assessment of the disease progression and curative effect can help predict the patients with severe diseases and provide more timely and accurate treatment for the patients with new crown pneumonia.
According to Lei Feng net (official account: Lei Feng net), the system is widely applied not only in Wuhan, Hubei, but also in the national medical institutions, and is the first time to transmit the experience and ability of China’s war epidemic to Japan, Italy, the United States, Germany and other countries through AI, and help the world fight against the epidemic situation.
After more than four years of development, the supposed AI products have been granted EU CE access and Japan PMDA access. However, the global epidemic situation is still continuing, and the global market value of this product will be further highlighted.
The second is the inferscholar AI academic research platform.
It is a software and hardware integrated device, providing zero code, whole process deep learning and imaging model research and incubation tools, providing a full stack of intelligent research infrastructure for medical researchers.
According to Lei feng.com, this platform uses full interface operation to create a visual big data analysis tool for medical researchers, and endows medical researchers with the ability to carry out high-level scientific research without computer programming ability. It can process all-round image data of X-ray, CT, Mr, pet, ultrasound and pathology. The research scope covers the brain, chest, abdomen, whole body bone tissue and other organs of the disease image intelligent analysis, especially for the disease early diagnosis, treatment monitoring, prognosis prediction of imaging artificial intelligence research has important value.
At present, the platform has been applied in many top hospitals in China.
Chen Kuan said that scientific research originates from clinical practice and is applied to clinical practice. The medical research platform infrastructure based on artificial intelligence technology is a part of the new medical infrastructure, which will play a more important role and value under the drive of the new infrastructure.
It is worth noting that on June 9, tui Xiang technology released a brand new and upgraded brand image.
Chen Kuan said that the move is to conform to the development trend of “new infrastructure”, create a brand-new medical AI infrastructure brand with the impression of speculation, and form an international medical AI brand.
He said that in the future, it is supposed that medical AI will effectively improve the ability of basic medical screening, diagnosis and treatment in the national new infrastructure strategic path; at the same time, it will continue to promote the development of Pan AI precision medical application fields with multiple diseases, multi departments, multi scenes and multi demands based on the excavation of clinical pathway for single disease.
Huiyi Huiying: three product systems to open up the closed loop of medical process
Chai Xiangfei, CEO of Huiyi Huiying, said that the new medical infrastructure needs to complete the construction of the overall project from regional imaging system, digital intelligent film, intelligent auxiliary diagnosis, remote imaging diagnosis, medical big data decision support system, interface development services, platform support, etc., to reflect the practical application value of smart medical.
On the whole, medical care is a heavy regulatory industry, both the use and charging links are related to the supervision of the higher authorities.
Whether it is drugs, devices, or artificial intelligence, wearable, 3D printing and other new technologies in the medical field, it needs to go through five stages.
First, it is necessary to form an expert consensus and get the recognition of experts in the industry; second, to rely on experts to form expert guidance; third, to obtain the approval of the drug administration; fourth, to approve prices, so as to charge hospitals; fifth, on the basis of deep implementation, to enter the scope of medical insurance, so that the application service can be changed from self payment to public reimbursement.
Chai Xiangfei said to Lei feng.com that the intelligent medical industry has formed a good expert consensus in 2018, and this year it may break through the third stage, that is, the FDA’s approval of this link, including overseas countries also attach great importance to access qualifications.
Recently, Huiyi Huiying multi disease imaging AI medical products have obtained the EU CE certification. The system includes lung CT image auxiliary diagnosis system, fracture auxiliary diagnosis system, chest X-ray auxiliary diagnosis system, etc.
He added that in China, artificial intelligence has been included in the guidelines for the treatment of multiple diseases this year, and the next step is to form a separate charging item, that is, to obtain the price permission before a large-scale commercial process can be officially started. The new infrastructure related policies will promote this process and accelerate the pace of medical AI landing.
According to Lei feng.com, at present, Huiyi Huiying has built novacloudtm intelligent image cloud platform, Dr The three product systems of turingtm Artificial Intelligence Aided Diagnosis Platform and radcloudtm big data artificial intelligence research platform have created digital, mobile and intelligent medical imaging, completed the closed-loop of screening, diagnosis and treatment decision support, and realized the whole process coverage from scientific research to clinical practice with data driven.
Up to now, Huiyi Huiying products and services have been applied in more than 1000 medical institutions, including more than 300 tertiary hospitals. In the field of medical imaging artificial intelligence research, Huiyi Huiying has undertaken and participated in more than 20 national and provincial scientific research projects.
In 2019, Huiyi Huiying carried out strategic upgrading and launched the overall solution of smart imaging, and continued to invest and layout in smart medicine, smart hospital, intelligent scientific research, big data and artificial intelligence applications.
Chai Xiangfei said that next, Huiyi Huiying will closely follow the idea of “new infrastructure”, continue to increase technology research and development and market investment, continuously improve the full volume and full cycle product solution of medical imaging artificial intelligence, which has opened up information flow, workflow and data flow, accelerate the landing of hospitals, assist hospitals and governments to establish an intelligent platform for the whole process of imaging medicine, so as to provide more mature services for customers Reliable technical and service support;
At the same time, we will continue to deepen the in-depth cooperation with Huawei, Intel and other partners to jointly build a brand-new scene of smart imaging medical services, and realize interconnected and healthy China.
Voxel Technology: “whole disease” idea runs through
Ding Xiaowei, CEO of voxel technology, believes that the construction of new medical infrastructure can start from encouraging medical informatization, Internet medicine, artificial intelligence + medical treatment, among which medical informatization is the infrastructure of modern medical industry. In this epidemic situation, Internet medicine and artificial intelligence + medical industry play an important role, including artificial intelligence CT film reading, online consultation and grading Diagnosis and treatment, telemedicine and other services.
“The new medical infrastructure can introduce these technology driven tools into the medical field, providing digital and intelligent support for the basic elements of the medical field, including hospital facilities, doctors, and medical equipment.”
The promotion of new medical infrastructure, especially the introduction of 5g, artificial intelligence and data center technology, will have a great impact on the medical AI industry. Among them, the construction of big data center, digital transformation and intelligent upgrading can bring more data and more powerful computing power.
Artificial intelligence itself is a part of the new infrastructure, and many project opportunities will emerge through the development; in addition, the connection, landing and transformation after the infrastructure construction can make various products more organically integrated into the diagnosis and treatment work.
In this epidemic, several products of voxel technology all played an important role.
When the epidemic broke out in February, voxel’s R & D team upgraded the whole disease screening solution of chest CT imaging to the special version of pneumonia, and entered Shanghai Ruijin Hospital and the first people’s Hospital at the first time to provide auxiliary film reading services for doctors, screen out suspected pneumonia patients, and prompt doctors to pay more attention to observation, which greatly reduced the work pressure of doctors.
In addition, voxel skin self-examination product “voxel skin knowledge collection” has played a huge role in the internet medical platform. On Tencent health home page, more than 200000 users use the small program online to check skin problems every day, and tens of millions of people have experienced it so far.
In addition, voxel’s ophthalmic disease screening products have been launched in more than 300 MMCs in China, and have also been launched in township health centers throughout the country, so that the people at the grassroots level can use the latest artificial intelligence tools, and make contributions to the popularization of eye health medical knowledge at the grassroots level.
Ding Xiaowei told Lei Feng that in the future, voxel will continue to make efforts in medical informatization, Internet medicine and artificial intelligence medicine, continue to deepen technology research and development and product optimization, realize more flexible deployment methods and enter more hospitals.
At the same time, we will continue to improve the integration of products online and downstream, create the whole process from intelligent devices to auxiliary diagnosis platform services to personal health management, collect and obtain more medical clinical evidence, and verify the efficiency improvement after construction.