On August 5, 2022, the shortlisted affiliations of "Innovative Task List of Artificial Intelligence Medical Devices" organized by the Ministry of Industry and Information Technology and Department of Medical Device Registration of National Medical Products Administration were officially released. The engineering medicine project entitled "Artificial Intelligence-assisted Diagnostic System for Cytological Screening of Endometrial Cancer" of Professor Li Qiling from Department of Gynecology and Obstetrics of our hospital became one of the 70 shortlisted projects in the direction of "Intelligence-assisted Diagnostic Products", ranking the 17th. It was the only shortlisted project from Shaanxi province, signifying that the innovative breakthrough of this project in the screening and artificial intelligence-assisted diagnosis of endometrial cancer has been recognized by national experts.
In 2012, Professor Li Qiling led the team to carry out a series of research and exploration on the screening and clinical screening sampling of endometrial cancer. In 2015, the team developed an accurate, minimally invasive and safe sampling system (Figure 1), including Li Brush, a disposable inverted cone-shaped double-sleeve intrauterine cell collection brush. The sampling success rate achieved up to 96.15%, which successfully resolved the bottleneck problemof clinical sampling.
Figure 1. Sample collection system
Figure 2. Pathological section system
Regarding the bottleneck problem of "difficult diagnosis" by clinical pathologists in cytological screening of endometrial cancer, Professor Li Qiling attempted to solve clinical issues by adhering to the concept of multi-disciplinary integration of science, technology and medicine. In 2018, the team cooperated with Professor Shi Guizhi’s team from Institute of Biophysics, Chinese Academy of Sciences and Professor Zhong Dexing’s team from Institute of Automatic Control and Detection Technology of Xi 'an Jiaotong University (XJTU) to jointly develop a pathological section system (Figure 2) and an artificial intelligence-assisted diagnostic system for endometrial cells (Figure 3). Serial studies have confirmed that two results can be obtained during one cycle of sampling, which reduces the difficulty of pathological diagnosis. The sensitivity, specificity and consistency rate of this system reach up to 92.0%, 92.2% and 93.5%, respectively. Besides initially solving the clinical problems of pathological diagnosis for endometrial cells, artificial intelligence can also assist physicians to identify and diagnose pathological sections in high-degree automation pattern, which can significantly relieve the workload of physicians and improve the work efficiency of pathologists. This system is applicable for large-scale screening and diagnosis.
Figure 3. Artificial intelligence-assisted diagnostic system for cytological screening of endometrial cancer
Adhering to the concept of engineering medicine, the team achieved hospital-enterprise cooperation and engineering medicine translation by utilizing the scientific and technological innovation academic exchange platform of engineering medicine and the research platform of Surgical Dreamworks. In 2021, the team signed a patent transfer and industrial translation agreement with Xi 'an Meijiajia Medical Technology Co., Ltd. regarding the core patents of this project. At present, the team has obtained 1 Class II registration certificate for scientific research project and 11 Class I filing certificates, and successfully put on market sale.
Since the establishment of this project, the team has carried out 12 clinical trials, published 11 articles in SCI-indexed journals, obtained 4 patents and 1 software copyright registration certificate, drafted 1 monograph, and hosted Medical Meeting of Endometrial Lesions and Pathological Progression for 7 consecutive sessions, aiming to promote the screening and pathological diagnosis of endometrial cancer throughout China.
