In today's era of big data and intelligence, the tradition of relying solely on the naked eye to detect cell morphology will soon be broken! On November 17th, it was learned from the launch conference of the collaborative research on artificial intelligence cell analysis in Chongqing that the artificial intelligence cell morphology analysis system developed by Chongqing Defang Information Technology Co., Ltd. is conducting further research and application in collaboration with top-tier hospitals such as Xinqiao Hospital, Southwest Hospital, the First Affiliated Hospital of Chongqing Medical University, and the Second Affiliated Hospital of Chongqing Medical University. This software, which brings together the most powerful "brain-like" experts and professors in this field, can increase the detection speed by about ten times while ensuring accuracy through intelligent scanning and analysis, and effectively reduce the detection cost.

If you go to the hospital when you are not feeling well, sometimes the doctor will ask you to have your blood, urine, stool and so on tested. Usually, these examinations are conducted by the naked eye with the aid of a microscope through empirical identification of the morphological and quantitative changes of the corresponding cells, and then a diagnostic conclusion is given. This conclusion is usually related to the experience of the examiner! Li Chunlin, the chairman of the German company, told the reporter that this is an empirical discipline and there are too few people with the same level of experience. Take Chongqing as an example. No more than ten people can understand and accurately distinguish it, because it takes at least ten years to train a morphology teacher before they can be qualified. Moreover, some tests take two or three hours to get results, and for patients, waiting is the most agonizing.

How can we analyze the cell morphology more quickly and accurately and draw scientific conclusions? The German company has applied artificial intelligence to the recognition of medical cell images, allowing AI to replace doctors in organizing patients' microscopic examination films. Doctors only need to review the test reports. Tests that previously took tens of minutes to complete can now yield results in as fast as 3 minutes, and the repeatability can reach over 98%.

To achieve such intelligent analysis, it must be based on big data. At present, the German company is collaborating with major hospitals in Chongqing (including hospitals of the Army Medical System, hospitals of the Chongqing Medical System, local hospitals, etc.) to uniformly collect and identify human cell data, formulate industry standards, and conduct research and application on human cell morphology based on artificial intelligence. After collecting massive amounts of data, the experience and analysis of experts in the field of cell morphology are input into the computer to form data diagnosis. "It's like equipping a computer with a 'brain-like' system," Li Chunlin said. This "brain-like" system integrates all the experience and wisdom of experts in this field. In the future, as long as patient cells are collected, through intelligent scanning, the morphological traces of related diseases can be quickly found, replacing the naked eye to obtain precise results.

Based on the breakthroughs in artificial intelligence in cell morphology, Chongqing Defang Company is currently researching the establishment of a human cell database to form the Defang Medical Cloud. At present, the Defang Medical Cloud has been initially established. At that time, even in primary medical institutions without experts or at home, as long as patients upload cell smears, including urine, feces, etc. to the cloud of Defang Medical, they can obtain accurate test results just like in large hospitals.

Li Chunlin stated that the significance of establishing a human cell bank lies in building a database of normal and abnormal cells from different parts of the human body, providing precise clues for the early detection and diagnosis of diseases related to cell morphological lesions, and also significantly reducing the cost of testing. (Source: Upstream News · Chongqing Business Daily