The Arterial Network was informed for the first time that Subtle Medical, a medical AI company focused on medical image enhancement, recently received approximately $5 million in Pre-A round financing.
The leading investors in this round are the top US venture capital firm Bessemer Venture Partner and seed round investors Data CollecTIve, Breyer Capital and Fusion Fund, etc. Angel Investment Fund, Baidu Venture Capital, Qingyuan Venture Capital, Wisemont Capital Continue to support.
The time to get the financing was only two months after Subtle Medical won the NVIDIA Startup Acceleration Challenge.
A photo of the Subtle Medical team and Huang Renxun
Subtle Medical is committed to making the image inspection process more efficient, safer and smarter. The company uses deep learning algorithms to improve the quality and diagnostic value of medical images, reducing the time, risk and cost of medical imaging. Subtle Medical has achieved remarkable results in large imaging fields such as MRI and PET.
Suben Medical founder Gong Enhao told the arterial network reporter that this round of funds mainly has three purposes:
First, the funds will be used for approval by the US FDA, China Food and Drug Administration, and European CE. The company's products are expected to be FDA-approved by the end of this year.
Second, expand the AI ​​image development team and increase the scope of clinical product cooperation.
Third, complete the layout of the medical industry chain. Subtle Medical will cooperate with more medical institutions and manufacturers around the world on the basis of more than a dozen top medical institutions in the United States, and gradually expand its business in Europe and China.
Medical imaging inspection in the United States is costly and inefficientCBInsights' latest medical consumption report points out that US medical spending totals about $3 trillion, accounting for more than 17% of GDP. Among them, medical images account for 10% of the total medical expenses. There is a lot of room for improvement behind the high medical expenses.
Gong Enhao told the arterial network reporter that in the United States, a nuclear magnetic resonance (MRI) charge is between $1,000 and $2,000, while a neutron tomography (PET) charge is in the thousands or even tens of thousands of dollars. Only about 10% of this cost is used to pay for imaging doctors' diagnosis. The remaining 80%-90% of the cost comes from the cost of purchasing the equipment, maintenance costs and the cost of the entire image inspection operation. In addition to the high cost, a longer appointment waiting time is not only inconvenient for the patient but also may delay the examination of the disease.
Therefore, improving efficiency is important for both hospitals and patients without compromising the quality of the diagnosis.
Since the conventional method of shortening the imaging time severely degrades the image quality, affects the accuracy of the diagnosis, and the problem of long image inspection time has not been completely solved for many years.
Gong Enhao told reporters that Stanford University developed compressed sensing technology to improve the efficiency of MRI examination more than 10 years ago. At present, major medical imaging equipment companies have their own similar technologies and many have entered the medical market through FDA certification. However, in practical applications, doctors do not fully accept this technique based on fixed models and statistical algorithms. The main reason is that the advantages of the algorithm in image quality and computational efficiency cannot meet the clinical needs.
Interdisciplinary team deepens scientific research and clinical transformationBridging the transformation gap between scientific research technology and clinical application requires not only professional technical development capabilities, but also awareness of clinical pain points. The core members of the Subtle Medical team have a dual background in medicine and engineering.
The founder of the company, Gong Enhao, graduated from Tsinghua University in 2012 with a bachelor's degree in biomedical engineering. During his time at the university, he worked with Philips scientists to optimize compression sensing technology and published several papers and patents. In 2012, he entered the Department of Electronic Engineering at Stanford University to pursue a Ph.D., and further explored research in the field of nuclear magnetic resonance. During this period, he realized that the clinical needs and market space in this field are very large, and that the research projects are out of touch with the actual clinical application.
Since the beginning of 2016, Gong Enhao has used his deep foundation in deep learning and artificial intelligence for several years at Stanford University to begin research on medical image post-processing, image reconstruction and auxiliary diagnosis using deep learning techniques.
In July 2017, Gong Enhao officially registered the company with Subtle Medical, together with Dr. Greg Zaharchuk, professor of radiology at Stanford University School of Medicine, neuroimaging doctor and director of the Frontal Neuroimaging Laboratory. Subsequently, two Taohua biomedical alumni Zhang Tao and Zhu Liren who were engaged in medical imaging research in the United States also joined.
Dr. Zhang Tao previously developed magnetic resonance imaging technology at GE Healthcare General Medical and served as a faculty member at MD Anderson Cancer Hospital. Dr. Zhu Liren has many years of academic and industrial experience in multimodal imaging.
As a graduate of biomedical engineering, Gong Enhao said that this profession has a strong combination of medical and industrial characteristics. They must use both engineering knowledge and basic medical knowledge, as well as clinical needs. Engineering techniques are used to solve clinical problems, so their professional development is very suitable for the research and work of the medical artificial intelligence industry.
The founding teams of many medical artificial intelligence companies at home and abroad have a professional background in biomedical engineering, and many of them are Tsinghua biomedical alumni.
Improve image efficiency and reduce radiationSubtle Medical's first product uses AI to improve the quality and efficiency of medical imaging diagnostics, empowering hospitals and imaging centers. As a team of AI experts and imaging doctors, Subtle Medical does not aim to replace medical imaging physicians, but focuses on the pain points of actual productivity. The company's starting point is to make image inspections more efficient, more economical, safer and smarter. Subtle Medical provides an AI image processing platform for hospitals and imaging centers, with a number of features in the first product.
Improve imaging speed and improve image quality: The product uses AI to improve the image quality of MRI magnetic resonance and PET neutron tomographic images, thus achieving 2-4 times MRI acceleration and 4-10 times PET based on existing imaging equipment hardware and software. accelerate.
Gong Enhao stressed that at this stage, Subtle Medical mainly deals with images of MRI and PET (including PET-CT and PET-MR). In these areas, Gong Enhao's laboratory at Stanford has decades of experience in research and industrial cooperation, with deep data and experience. At the same time, from the perspective of clinical needs, the speed of MRI and PET imaging is the slowest, and the pain points of doctors and patients are also the most obvious. MRI can see more soft tissue and contrast. Both PET-CT and PET-MR provide molecular and functional imaging and are of great value in many clinical tests, such as early cancer screening and staging.
Reduced radiation dose and reduced inspection time: While improving image imaging speed and quality, Subtle Medical is also using the power of technology to reduce the use of radiation doses, reducing at least 10 times the contrast agent and 100 times the radiation hazard. Both PET-CT, PET-MRI and PET require the use of radioactive agents, which are at risk of radiation, especially PET-CT, which has a radiation dose that is several times that of a head CT. Gong Enhao said that their recent papers mentioned that their Stanford laboratory can reduce the use of radiation doses by 100 times-200 times under the premise of ensuring image quality. In clinical practical applications, the goal will be at least The radiation dose is reduced by one tenth of the original while ensuring image quality.
At the same time, Subtle Medical's system can be used simultaneously to increase the speed of image inspection and reduce radiation dose.
Product embedding doctor workflowGong Enhao said that while they are developing product quality, they will also consider the doctor's workflow and will not burden the doctor. Subtle Medical applies AI technology to the forefront of image inspection. After the image data comes out of the device, it enters their system directly. After processing, they enter the PACS imaging workstation. The doctor sees the system processed image, and the workflow is the same as before, which does not burden the doctor.
Because Subtle Medical's system is applied to the forefront of image inspection, its system can get first-hand medical image data. Their partners can be medical imaging device manufacturers, but PACS system manufacturers and medical AI companies. For the three parties, it is very attractive to ensure even enhance the image quality and reduce the radiation dose without increasing the doctor's workflow.
Gong Enhao said that the company's first FDA (Class 2 510k) product is expected to pass in October-November 2018. Subsequent follow-up will continue to follow up on multiple FDA applications and continue to expand the product line. Taking the optimized image inspection process as a breakthrough, the image AI platform was built to optimize the image acquisition and analysis process.
Gong Enhao told reporters that the reason for the certification of the second type of certificate is because Subtle Medical aims to reduce the cost of imaging examinations and shorten the time of examination by enhancing the quality of medical images, thus helping doctors to make diagnoses. The specific indicators can be quantitatively judged, and the product itself does not directly give the diagnosis result, so there is no need to declare the three types of certification.
Stanford's millions of high-quality medical image data supportThe use of deep learning technology to develop models is inseparable from the support of high-quality big data, Subtle Medical is no exception. According to Gong Enhao, since they were born in Stanford, Stanford authorized the use of three patents and massive medical image data (a rough estimate of more than two million medical images recently used, mainly MRI and PET/CT, PET/MR images) Patents and data are the result of their team's research at Stanford.
In addition to the data accumulated by these Stanford researchers in their usual research, Subtle Medical is also working with leading hospitals and third-party imaging centers in the United States to use clinical data from these institutions while ensuring privacy and compliance.
It is worth mentioning that Subtle Medical does not hire a large number of doctors' teams to label these data, but to design, collect and process specific medical image data according to the research needs, and convert it into the data needed for research and development.
Gong Enhao said that it takes many hours or more to collect a complete medical image data. Only in the scientific research environment can high quality be achieved, and the image background of researchers is also very high.
Such a large amount of data is not available overnight, and has been accumulated by researchers at Stanford for many years. The quality of these data has also been verified by research for many years.
Talking about the diseases involved. Gong Enhao said that Subtle Medical's research focuses on modal research rather than specific disease research, and the product is focused on image enhancement and post-processing. Any disease that uses MRI, PET-CT, PET-MRI, PET and other radiological equipment for routine examination, such as stroke, brain tumor, lung cancer, Alzheimer's disease, etc., can provide image enhancement services.
Accelerate market layout and product landing and cooperationCurrently, Subtle Medical's international market road has begun.
Subtle Medical is based in the US market and has partnered with more than a dozen leading medical schools, hospitals and third-party imaging centers in the United States for clinical and clinical system testing. Partners include Stanford, UCSF (University of California, San Francisco School of Medicine), MD Anderson (MD Anderson Cancer Hospital), Mayo Clinics, OHSU (Oregon Health and Technology University), Hoag Hospital (Hogg Hospital) And the largest image center in the United States, RadNet and other institutions.
In the industry, Subtle Medical has also actively cooperated with medical imaging and artificial intelligence companies such as Neusoft Medical and NVIDIA to optimize medical image processing technology based on artificial intelligence.
This round of financing will also accelerate the market layout and product landing.
Outdoor Fixed LED Display is a popular product for its high quality, every year sold to at least 80,000 pieces around the world, including Europe, North America, southeast Asia.Compared to other indoor LED display in the market, its biggest advantage is that it can display high-definition images while maintaining low power consumption.Besides, it adopts Die casting aluminum cabinet which is ultra-thin and ultra-light and owns good heat dissipation.Easy to install and maintain and suitable for multiple indoor scenes.
Application:
* Business Organizations:
Supermarket, large-scale shopping malls, star-rated hotels, travel agencies
* Financial Organizations:
Banks, insurance companies, post offices, hospital, schools
* Public Places:
Subway, airports, stations, parks, exhibition halls, stadiums, museums, commercial buildings, meeting rooms
* Entertainments:
Movie theaters, clubs, stages.
Outdoor Fixed LED Display,Led Wall Display Screen,Curved Led Display Screen,Led Display Board
Guangzhou Chengwen Photoelectric Technology co.,ltd , https://www.cwstagelight.com