In the field of medicine, pathological diagnosis is the final diagnosis of a disease. It plays a vital role in defining the nature of the disease, explaining clinical symptoms, determining the principles of treatment, assessing prognosis, and determining the cause of death, and is the gold standard for disease diagnosis.
Pathologists as "doctors of doctors", due to the long training cycle, has long led to a huge shortage of pathologists in the mainland. According to a 2019 survey by the National Health Commission's Medical Administration Bureau, the number of pathologists registered in the mainland is about 12,000. According to the requirements of the National Health Commission, if every 100 beds are equipped with 1-2 pathologists. At present, the shortage of pathologists has exceeded 100,000.
With the rapid development of modern science and technology, pathology has also made great progress. However, due to the imbalance in geographical development, some pathology equipment is outdated, and the diagnostic level and diagnostic standards of pathologists are not uniform, which seriously affects the efficiency and accuracy of clinical pathological diagnosis. Therefore, the development and realization of pathological artificial intelligence auxiliary diagnosis, digitization and intelligence has become an inevitable trend in the development of pathological modernization.
Since 2017, a boom of "medical + artificial intelligence" has emerged in China, and nearly 100 pathological AI companies have appeared in the industry. However, the professional threshold of pathology is extremely high, and it does not have the soil for the development of the "short and fast" model of the consumer Internet industry. In the past 5 years, the domestic pathology AI industry has experienced a big wave of sand, according to incomplete statistics, the number of enterprises focusing on pathological AI is only about 10.
Guangzhou Kunyuan Fangqing Medical Technology Co., Ltd. (hereinafter referred to as "Kunyuan Fangqing"), a technology company that "deeply cultivates pathology and embraces intelligence". As the only intelligent pathology R&D team in China pioneered by pathologists, its strategic goal is to promote the comprehensive and intelligent upgrading of "production, learning, research and use" in the pathology industry, and has the ability to continuously innovate R&D and professional services in the pathology industry. Therefore, in addition to the multi-disease AI auxiliary diagnosis, the product layout of Kunyuan Fangqing also includes AI-assisted quality control and department management, AI-assisted pathology research platform, etc., and has accumulated 16 patents and 19 soft works, which has developed rapidly in the industry.
Difficult to replicate expert R&D team,
Build high barriers to competition
As one of the early enterprises to enter the field of pathological AI, the steady development of Kunyuan Fangqing is due to its "pathology expert-led" team.
In 2018, Professor Ding Yanqing, the "Nanding" of the Chinese pathology community, founded Kun Yuan Fangqing and served as the company's chief scientist.
Professor Dinghiko Ao
Professor Ding Yanqing is a well-known pathologist in mainland China, the president of the 3rd Chinese Association of Pathologists, a national-level teaching teacher, and enjoys special allowances from the State Council. He is also the world's first autopsy of the remains of a SARS patient, which confirmed for the first time that SARS is a viral infection; from 2020 to 2021, Academician Bian Xiuwu and Professor Ding Yanqing were selected as "Elsevier China Highly Cited Scholars", and are currently the only two Chinese pathologists on the list for 2 consecutive years.
After practicing medicine and teaching for more than 50 years, Professor Ding Yanqing has trained a large number of talents including "National Jieqing" for the pathology industry in the mainland. It also has a forward-looking strategic vision and boldness in the combination of pathology and new technologies. As early as 2010, he led the remote diagnosis pathology project cooperated with Huayinkang Group and created a pathology digitalization and remote co-construction model of growth services, and the commercialization progressed smoothly.
In 2016, Professor Ding Yanqing once again gained insight into the new trend of intelligent development of the pathology industry, and subsequently founded an artificial intelligence research and development company based on pathologists, "Kun Yuan Fangqing", to help the intelligent development of pathology diagnosis in mainland China.
The core research team of "Kun Yuan Fangqing" is led by pathologists, and at the same time absorbs technology development talents from universities and research institutions at home and abroad. According to Mr. Shang Bin, CEO of Kun Yuan Fangqing, there are currently more than 30 pathologists and 5 full-time pathologists in the R&D team. As the only pathology AI research and development team in China led by pathologists, Kun Yuan Fangqing has also built its own strong competitiveness.
From scientist founder to pathologist and algorithm engineer, the large age span is another feature of Kun Yuanfangqing's team. Speaking of this feature, Shang Bin said: From the perspective of actual results, the complementary and multidisciplinary integration of the team shows strong development potential.
Intelligent assisted pathological diagnosis research as a hot spot in the field of pathology. How to use AI to solve pathological diagnostic pain points? How to turn advanced theories into reality to promote the development of the industry? Knowing the problems and needs is the first priority.
A brief analysis of the advantages of "pathologist-led"
Shang Bin said: "The industry barrier of pathological AI is that it is highly specialized, and data is the biggest bottleneck. The accuracy of AI depends on the ability to provide a large number of accurate pathology annotation data, and the refinement of these data still relies on the deep involvement of senior pathologists. ”
The industry bottleneck that hinders the industrialization of pathological AI has become a natural advantage of Kunyuan Fangqing. Kunyuan Fangqing gave full play to this unique core advantage, and pathologists participated in the whole process of product research and development, such as selection points, options, data labeling, algorithm research and development, tracking quality control, etc., to ensure that in the process of innovation and research and development, error correction and risk elimination were eliminated in a timely manner.
Incubated the first large-scale clinical cell pathology AI,
The results have attracted widespread attention
In 2021, Kunyuan Fangqing cervical cytology AI-assisted diagnostic products were approved for marketing by the State Food and Drug Administration. The system was validated by more than 1.06 million clinical samples, and the sensitivity of AI diagnosis > 99.9%, and the specificity > 80%. At present, it has been applied to more than 70 medical institutions, becoming the first pathological AI auxiliary diagnosis platform in mainland China that is truly applied to the clinic on a large scale.
A large number of verifications have shown that the system has high diagnostic accuracy and can identify clinically difficult and rare lesion categories, and has achieved a number of innovations. The research results of the project were published in the Journal of Chinese Pathology and Nature Communications respectively, which attracted wide attention from the mainland and global academic circles.
As one of the "two cancer screening" programs, cervical cancer is a common gynecological malignancy. In recent years, its morbidity and mortality rate have gradually increased, and it has shown a trend of rejuvenation. The "Healthy China Action" (2019-2030) aims to improve the coverage of "two cancers" screening, and clearly proposes the introduction of artificial intelligence-assisted diagnostic technology. With the support of national policies, Kunyuan Fangqing's innovative products can be said to have caught up with the "hitchhiking" of industrial landing.
Up to now, Kun Yuanfangqing's cervical cytology pathology pathology AI system has been successfully approved and deployed in more than 70 medical institutions, including large-scale top three hospitals, as well as a smart pathology diagnosis laboratory platform with a peak business of 5,000 cases/day.
According to reports, in addition to improving the accuracy of interpretation, the system can greatly improve the work efficiency of pathologists: first, the system works without interruption, and the minimum configuration can complete 1,000 sample analysis work per day; second, the system automatically labels suspected diseased cells, assists pathologists to locate the location of suspected diseased cells, quickly completes the review, and the doctor's diagnosis time is increased from 3 minutes to about 15 seconds; third, the system can directly generate a report after the doctor reviews and signs in accordance with international TBS standards. Realize the automation, intelligence and informatization of pathological diagnosis.
Sustainable R&D capabilities are key,
Wisdom pathology is not limited to the aids of single-disease AI
Kunyuan Fangqing's product strategy closely fits the new trend of the development of "next-generation diagnostic pathology" in the industry, based on pathological morphology and clinical information, combined with molecular detection, bioinformatics analysis, artificial intelligence and other technologies, multi-omics, cross-scale comprehensive diagnosis.
How to let pathologists get rid of repetitive work and devote more energy to the identification of difficult lesions; how to comprehensively build an intelligent pathology department in combination with clinical needs, and accelerate the construction of departments and talent training are important considerations for the research and development control of Kunyuan Fangqing.
Its product research and development ideas are: in the AI-assisted pathology diagnosis project, take the lead in breaking through the project with high diagnostic quality, repeatability and feasibility, and then exert efforts on projects with high complexity and difficult diagnosis; in the construction of AI auxiliary departments, according to clinical professional specifications, develop AI-assisted quality control and management systems, AI-assisted pathology research platforms, etc. Thus forming a product echelon with easy entry and difficulty, from simple to complex.
Source: Kun Yuan Fang Qing
Kun Yuan Fangqing has completed the product research and development of cervical fluid-based cytology, serous effusion cytology, thyroid aspiration cytology, lung fibromytomylotomy and cytology AI-assisted diagnosis; completed the cytology of cytology; the construction of a precise judgment system of cytopathology AI and the accurate judgment system of multi-cancer immunohistochemistry; and the platform of multimodal fusion and multi-disease pathology AI assisted diagnosis is under development. It is expected that by the end of 2022, the new project will be approved for medical device registration certificate.
In addition, AI projects such as prostate cancer phenotype prediction gene changes and prognosis analysis models, customized segmentation of multi-cancer lesions, and special types of structural detection (such as cancer thrombosis detection and vascular invasion detection) are also being promoted.
Accumulated 80 million financing,
Accelerate capacity expansion and sales channel expansion
According to reports, Kun Yuanfangqing has established cooperation with scientific research institutions and medical institutions such as the Institute of Automation of the Chinese Academy of Sciences, the Southern Hospital of Southern Medical University, the Southwest Hospital of the Army Military Medical University, the relevant hospitals in Shenzhen, and dozens of institutions such as Huayinkang Inspection Group. And has completed two rounds of financing, a total of 80 million yuan in financing amount. The company is currently conducting the third round of financing, mainly for new AI projects and related equipment expansion, marketing and sales team and channel establishment.
At the same time, Kunyuan Fangqing is building an overall solution for smart pathology, laying out equipment such as digital pathology slices and related consumables in the upstream of the industrial chain, as well as professional smart pathology laboratories that provide downstream services, and intends to establish AI pathology auxiliary diagnosis centers in various provinces and cities. At present, the company has obtained some software, equipment, reagent consumables of medical equipment certification.
Source: Kun Yuan Fang Qing
In the interview, ShangBin said: "With the State Food and Drug Administration gradually liberalizing the approval of artificial intelligence products, pathological AI will usher in an explosive development period in the foreseeable future. ”
"Focusing limited energy on the career of infinite possibilities" is a portrayal of Kun Yuanfang's cultivation of the pathological AI industry. In the future, Kun Yuan Fangqing will seize the opportunity of industry development, quickly introduce mature products to the market, and quickly enter the actual work of more pathologists. At the same time, we will continue to promote the development plan of new products, jointly promote Chinese pathological AI to the forefront of the world, and jointly create a new era of Chinese pathological artificial intelligence with the industry.