The deep learning algorithm based on convolutional neural network has gradually become a new kinetic energy of medical revolution since 2012. Intelligent assisted diagnosis and treatment systems based on medical images such as CT, MRI, X-ray, ultrasound, thermal infrared, cell smear, and electrocardiogram have been proven effective in clinical use.
Not long ago, the author was in a chat with Zou Guowen, a investment partner of Dow, and Mr. Zou revealed that the number of enterprises committed to developing intelligent auxiliary medical treatment systems in China exceeded 100. The fact that so many entrepreneurs or entrepreneurs have escaped from this has proved its great contribution to medical progress, and on the other hand, it indicates its potential and huge commercial value space.
The virtual reality behind the investment turmoil is basically clear in 2018. Some have failed to get financing, and the capital chain has broken. Some of them have been smashed in the form of “selling sheep and selling dog meatâ€. Some of them have been working hard for many years, and they are still working hard to make AI innovation. Some rely on technology, funds and medical resources. The leader of strength is still leading. The general market structure is basically stable this year, and only a few of them are left in every segment.
If in 2017 everyone is still working on the accuracy of algorithm and image recognition, it will be certified and commercialized in 2018. How to sell money for products will become the key research topic of this year's enterprises. Since last year, Yiou think tank has talked with many entrepreneurs and investors about related topics, and has formed some unconformed views and opinions. I hope to share and discuss with experts.
First, the current mainstream products in the market
The identification and diagnosis of pulmonary nodules is the starting point for the training of most enterprise algorithm models. On the one hand, because China is a large lung cancer country, the amount of lung image data is the most abundant; on the other hand, the global research on lung nodule recognition is the most mature. Domestically, the rapid progress in the diagnosis and screening of lung cancer, mainly including the technology, Tuoma Shenwei, Jianpei Technology, Voxel Technology, Yitu Medical, Huiyi Huiying, Shenrui Medical, Vision Medical, China Resources Wanli Cloud, zero-tech and so on. In addition to lung cancer, automatic detection of breast cancer, liver cancer, nasopharyngeal cancer, and stroke is also becoming a focus.
Intelligent CT-assisted screening product (AI-CT) function interface (Source: Imagine Technology)
The auxiliary screening of cervical cancer, namely the cervical smear intelligent assisted screening system, is an area that is attracting industry attention. This application is mainly for the pathology department, and the assistant pathologist completes the automatic detection of cellular DNA on the pathological section. Typical companies include Landing Medicine and DeepCare. The former core product cell DNA automatic detection analyzer is the first product in China to receive the second class certificate.
The diagnosis and screening of diabetic retinopathy is increasing. Voxel Technology, Airdoc, DeepCare and other companies have launched related products.
There are also a few applications that involve fewer companies:
· Tencent's hard-to-sell esophageal cancer screening products;
· AI-ECG-Platform, introduced by Lepu Medical, is an automatic analysis and diagnostic product for ECG;
· Fracture X-ray intelligent diagnosis system launched by Huiyi Huiying;
· Diagnostic screening of neonatal jaundice launched by Naobao is an AI product for skin images.
In addition to the diagnostic link, there are products worth looking forward to in the treatment chain. In the field of radiotherapy, companies such as Lianxin Medical, Huiyi Huiying, and Global Medical have launched automatic target delineation systems to help doctors complete radiotherapy programs efficiently and accurately.
In the field of surgery, Qijian Medical, China Resources Wanliyun and other companies have launched an image 3D reconstruction system based on evolutionary algorithms and VR technology to help doctors more accurately preoperative planning.
Second, the starting point is in the top three, and the end point is in the grassroots
The training of algorithmic models requires a large amount of high-quality, diverse image data, which is the most in the top three hospitals. Therefore, the company strives to cooperate with the top three hospitals through scientific research cooperation, and obtains high-quality data, which has become the basic route. However, the top three hospitals do not lack high-quality imaging and radiotherapy doctors, and primary hospitals are lacking. Therefore, from the perspective of practical application, the real demand is at the grassroots level and not at the top three. The medical informationization, grading diagnosis and treatment, and telemedicine that the country has promoted in the past two years are a big positive factor. Based on digital imaging and on the condition of instant interoperability of multi-level hospital medical records, it will enable grassroots hospitals to truly have the ability to diagnose and treat major diseases such as cancer.
Many companies build products in the cloud, from the private cloud of a hospital to the public cloud of many hospitals, which is the product form for future product applications.
Third, do you want to do hardware?
Most of the companies mentioned above started out as software systems. There are two kinds of profit models for software: one, sold to the hospital, and the annual fee charged to the hospital; second, cooperation with the hospital, according to the patient's head charges. But what about the financial models of these two profit models is still unknown. As several mainstream companies get certified and begin to enter the market stage, domestic medical device manufacturers are likely to extend the "Olive Branch" to seek cooperation or mergers and acquisitions. The combination of the two can quickly find the path of income generation by means of the resource strength of medical device manufacturers and the high value-added services of soft and hard integration. For medical device manufacturers, the added value of products can be higher. It can also enhance the technical strength and market competitiveness of enterprises.
Do you want to create your own hardware? This is a question that deserves careful consideration. The advantage of doing hardware is to make money from consumables. In the medical device market, what really makes money is not how many devices are sold, but consumables. Taking Lanting Medicine as an example, the income in 2017 is nearly 100 million. The products it relies on include the cell DNA automatic detection analyzer mentioned above, as well as cell fixatives, stains, disposable cervical brushes and other consumables. There are also many companies that do to the C gene sequencing service, and the profit point is mainly in the body fluid collector. Of course, doing hardware is relatively heavy for companies, and it is not in line with every situation, so it needs to be cautious.
In general, Yiou think tank believes that 2018 will be the starting point for the commercialization of medical AI enterprises. The certification progress of CNDA is also accelerating. There are already several companies that have obtained the second-class certificates. The dust settles. We are very optimistic about the tremendous improvement that artificial intelligence brings to the diagnosis and treatment of diseases in the next five to ten years. We also hope that we can enjoy the technological improvement and bring us healthy and quality life.
SHAOXING COLORBEE PLASTIC CO.,LTD , https://www.colorbeephoto.com