Medical AI no longer "telling stories", Heis heterogeneous comprehensive productization

For some time, the topic of AI bubble has received much attention. In the past few years, the "story" about AI has been heard. Today, society and the market seem to care more about one issue - can AI fall?

The medical field is a fierce battlefield for the AI ​​R&D team, but so far there has not been a AI artificial medical device that is truly legally listed, but this day may come soon.

The 2018 Chinese Conference on Digestive Endoscopy is at 8.26. In Nanjing, at the annual top event in the digestive endoscopy world, two digestive endoscopic AI reports were related to a heterogeneous medical AI R&D company, Hex's Heterogeneity; at the same time, Xi's R&D was also exhibited during the conference. A variety of medical devices and cloud application technologies related to digestive endoscopic imaging.

Different from the medical AI research peers, the Xi's demonstration, whether it is the conference report or the application equipment, is centered around the “clinical application”, which means that the company that has the ability to outperform the medical AI has its research and development results. In the stage of mass production and hardware equipment.

The Arterial Network reporter interviewed Song Jie, the founder of Heis. "As a company that has been deeply immersed in the medical AI field for three and a half years, at this stage, Xi's will not go to 'tell vision, tell stories', the market can accept the real application technology, it is the clinically acceptable to sell. Products, equipment." Song Jie said.

Song Jie gave a keynote speech at the 2018 China Digestive Endoscopy Annual Meeting

During the conference, the reporter learned that Xi's research and development results in the field of digestive endoscopy AI have covered the entire digestive tract, involving more than 100 species including tumors, polyps, atrophic lesions, ulcer lesions, erosive lesions, and vascular variability. At the same time, Hirsch will also introduce endoscopic AI early cancer recognition technology for a variety of different light sources in the coming months. Related application technologies have been exposed with a variety of medical equipment (digestive endoscopic AI real-time image judgement, capsule endoscopic image AI analysis and judgment instrument, etc.) and cloud service products, and form a three-dimensional AI product program in this field.

Xi's exhibited digestive endoscopic AI equipment (1) - digestive endoscopic AI real-time determinator

Histo digestive endoscopic AI products (2) - AI digestive endoscopy

Comprehensive layout of medical AI, multiple devices to declare Class III devices

In the field of digestive endoscopy AI, the industry has a lot of attention. In July 2017, the joint research and development results of Xi's and Huaxi Hospital were deeply reported by CCTV-1 News, Xinhua News Agency and other heavyweight media, and were evaluated as Leading the international market; the first international “AI Digestive Endoscope” launched in November, for the first time, realized the ultra-high-speed, hysteresis-free real-time video judgment during the digestive endoscopy process, and was once again concerned by the mainstream media and the industry. The display is also expected to advance, but the progress of research and development in other medical segments has not been disclosed.

According to Song Jie, Xi's has a good application-level technology and equipment products in CT, ultrasound, ECG and other fields, and asked about the details. Song Jie said that Xi's research and development entry points in these fields. Unlike most of its peers, Hirsch is more focused on clinical applicability.

In the first half of the year, Hirsch has begun to declare the application of Class III medical devices for multiple products, including three digestive endoscopic related products and three other medical professional products. The highlight is that related AI technologies have efficient hardware medical devices. Products, not just at the recognition software level.

Digestive endoscopy AI Hei's depth research and development

According to the 2015 cancer statistics published online by CA Cancer J Clin, the annual incidence of gastric cancer in China is as high as 679,100 cases and about 500,000 deaths. 17.7% of total cancer deaths.

As the most direct inspection equipment for digestive tract diseases, digestive endoscopy has become very popular in China, but the early diagnosis and treatment of digestive diseases, especially malignant tumors in China, has not been carried out very well. The reason is simple: lack of Experienced digestive endoscopy doctor.

An experienced doctor needs at least a training period of more than ten years. The cultivation of talents depends on the accumulation of time. No one can train enough medical talents for a short time. Then we will find an AI assistant for the old doctor. Find an AI teacher for a young doctor.

“With our existing medical system, it is impossible to complete the screening of digestive tract covering more than one billion people in China, but AI can empower us.” At this stage, we cannot expect AI to recognize diseases that we have not yet understood. But we know that AI can multiply our existing capabilities, which is what it means!"

"We will not stay in the research of a certain disease, we have to do the disease identification technology of the whole digestive tract; we provide a full range of products including equipment and services." Song Jie at the Chinese Digestive Endoscopy Annual Meeting put Hex's The idea is very thorough.

Xi's Digestive Endoscopy AI Equipment (3) - Capsule Endoscopic Image AI Analysis Determinator

Embedded in the "cloud" side AI service products to be in the word

Xi's has already developed the application of the cloud at the same time as the related AI technology hardware equipment. For the progress of this kind of service-oriented products, Song Jie said that “once the certificate is obtained, it can be used immediately”.

For such products, for primary medical institutions and medical examination centers, the cost of use is very low, and inspections that do not require high real-time requirements can be satisfied.

Now, Xi's is working with a number of top hospitals to establish quality control systems in digestive endoscopy, ultrasound, CT and other fields to help primary hospitals improve diagnostics and increase business income.

Xi's Digestive Endoscopic AI Service Product - "AI Digestion Cloud"

The medical AI research and development test is that technology is an understanding of the industry.

For the research and development of medical imaging AI technology, Xi's believes that this is first of all a cognitive problem for the direction of the industry. If there is no deep understanding and foresight of this industry, it is impossible to achieve much. If you do not understand the characteristics and application scenarios of each segment in the first place, it is definitely not feasible to "do the data and do research and development for research and development."

"For example, endoscopy, ultrasound inspection, solve the problem in real-time process, out of real-time, the application value is greatly reduced; our research and development of ECG AI is also based on real-time, because there is real-time Technology, for high-risk patients, can save lives, time is the first factor; for example, CT must first solve the problem of image quality, and at the same time, to find out more clinically valuable diseases, and Not to find a disease; there are many things that require deep thinking, but it requires industry experience and judgment."

"Development of medical imaging AI requires 'broadness' (multiple diseases) and 'deep' (special diseases); there is no 'depth' with breadth support, and the value at this stage is not great."

As a clinician, Song Jie, who has more than 20 years of experience in R&D and management of the pharmaceutical industry, naturally has his reasons.

In general, the company's philosophy is based on a deep understanding of the industry, and products based on this concept may be more promising.

From digestion to the beginning to multi-field flowering

Song Jie said: "Everyone knows that Xi's is engaged in digestive endoscopy AI, but we are not only digesting, the reason is very simple, Xi's has many years of research and development experience, there are many top hospitals for medical cooperation resources, there is a strong calculation At the same time, the commonality of medical image AI research and development in various disciplines is greater than the difference. Since we have experience, ability and resources, we will naturally do more subdivisions, 'more than one cloud, there will be one more chance of rain'. ”

It is reported that in addition to the establishment of the "Huaxi-Xi's Medical AI R&D Center" with the West China Hospital, Xi's has cooperated with many first-class hospitals in different fields. He has strong medical resources. Announced in June this year, Xi's own AI supercomputer, designed by high-performance computing expert Dr. Wu Ren, Shennong No. 1, provided a powerful calculation for Xi's.

Based on 64 of NVIDIA's latest Tesla V100 GPUs, the parallel computing power is over 90%, and it is also the most powerful medical image supercomputer in China.

The elements of medical AI research and development are just the following: industry background, medical resources, AI technology, computing power, industrialization ability, in the eyes of Xi's, medical AI research and development is complex and simple.

Shennong 1 (SINOSEEDS) medical image dedicated supercomputer

Looking into the future Heterogeneous planning

"As I said at the outset, Xi's does not need 'storytelling' at this stage, just want to show our real products, but we do have a bigger vision. In my eyes, from a business perspective, AI is just a 'skill weapon' to overcome the traditional medical system. His goal is to find a 'territory' for himself in the future under the new medical situation and lead future medical services."

AI is a weapon that can bring opportunities, not a simple technical product. ”

From the description of Song Jie, Xi's seems to have a bigger goal.

Indeed, the AI ​​field bubble will continue to be squeezed out, and the winner must not be a "story", but a real usable technology and a good product that can be seen and touched.

Yellow Waxy Corn Cob

Corn is a good source of vitamin C, which can help boost the immune system and help fight cancer. Compounds found in corn have been shown to reduce the risk of cardiovascular disease. According to research, fresh corn contains 4-5 times more body fat than rice and flour, and contains unsaturated fat, of which 50% are fatty acids, which inhibit the digestion and absorption of cholesterol. It is an excellent medicine for long-term use to reduce blood cholesterol and soften blood vessels, making it an ideal vegetable oil for patients with hypertension, coronary heart disease, obesity and the elderly.
Corn is also a cereal and low-fat food. Researchers have found that a low-fat diet has no significant effect on the risk of several diseases. In past follow-ups, a low-fat diet was found to provide significant and sustainable health benefits.

Yellow Waxy Corn Cob,Yellow Waxy Corn,Waxy Corn,Glutinous Corn Cob

Yomifresh , https://www.yomifresh.com