Medical enterprises collaborate with artificial intelligence to assist in early detection, with 555000 cases of colorectal cancer detected annually in China | Tencent | Artificial Intelligence
Through the use of artificial intelligence assisted detection products, the detection rate of polyps has been improved, and the incidence and mortality rate of colorectal cancer have been reduced. Recently, Tencent Seeking Pictures, in collaboration with multiple medical institutions, launched the "Electronic Endoscopic Image Assisted Detection Software for Colonic Polyps" to achieve real-time assistance and reminders for clinical doctors through artificial intelligence, which will effectively improve the early detection and intervention of diseases.
Data shows that the number of new cases of colorectal cancer in China has increased from 388000 in 2015 to 555000 in 2020. It is expected that the number of new cases and deaths from colorectal cancer will double by 2040. The early detection rate of colorectal cancer needs to be improved. Experts explain that the ability of colonoscopy examiners to detect adenomas is easily affected by their technical level, fatigue level, and alertness, and there are also significant differences with individual doctors. However, colonoscopy examination is cumbersome and invasive, which can easily lead to poor patient acceptance and compliance. The regional economic disparities and insufficient per capita medical resources in China make it difficult to ensure the quality of screening, which further restricts the promotion of early screening, diagnosis, and treatment of colorectal cancer.
The electronic endoscope assisted detection technology and products based on artificial intelligence can assist doctors in accurately and real-time detection of lesions, reduce missed diagnosis, and have extraordinary significance for grassroots medical institutions. As early as 2018, Tencent Maying collaborated with the National Digestive Disease Clinical Medical Research Center to promote the application of AI technology in colon tumor screening.
This research relies on the research project of "National Colorectal Polyp Management" led by the National Clinical Research Center for Digestive Diseases, led by Professor Li Zhaoshen, an academician of the CAE Member, and the director of the Department of Gastroenterology of Changhai Hospital of Shanghai Naval Medical University, and the deputy director Professor Bai Yu is responsible for the implementation, with the joint participation of several top three hospitals. After five years of significant development, the "Colon AI Solution" has emerged, leveraging Tencent's years of accumulated video parsing technology and image analysis advantages to help endoscopists draw a "hidden map" of the human digestive tract and accurately locate the coordinates of polyps. When improving the quality of hospital image annotation through multi center cooperation, it also effectively solves the two major bottleneck problems of the lack of effective training data for domestic lower gastrointestinal AI assisted diagnostic systems and the lack of unified industry standards.
The reporter learned that the "Electronic Endoscopic Image Assisted Detection Software for Colonic Polyps" needs to cooperate with colonoscopy examination. The video image signal is imported from the video stream output by the electronic endoscopic image processor, and after deep learning algorithm analysis, the suspected polyp position is marked in real-time in the video to remind doctors to pay attention to the suspected lesion. So far, the People's Hospital of Zhejiang Province and the Central Hospital of Wenzhou have used clinical trial pathways to evaluate that compared to conventional colonoscopy, colonoscopy assisted by this software can effectively improve the detection rate of polyps.
It is also reported that in the future, Tencent Seeking Film will reduce the industry threshold by building a common technology open platform and platformization of industry university research cooperation, establishing a technical standard system, sharing basic resource libraries, etc., promoting upstream and downstream cooperation in the industry, and promoting the application of medical imaging AI achievements from the laboratory to the bedside.
