91360 Secures China's First Class III AI Medical Device Certification for Cervical Cytology Diagnosis, Marking the Dawn of AI-Driven Pathology
Recently, independently developed by 91360 Medical Technology Nanjing Co., Ltd. (hereinafter referred to as "91360")“Computer-Aided Analysis Software for Digital Pathology Images of Cervical Cytology” has been approved by the National Medical Products Administration (NMPA), officially obtaining the first Class III medical device registration certificate for AI in the field of cervical cytology.(Registration Certificate No.: Guo Xie Zhu Zhun 20233210272). This signifies that the application of artificial intelligence in the field of cervical cytology has entered a phase of substantive implementation, and also marks a revolutionary breakthrough in precision-assisted diagnostic application scenarios within China’s pathology diagnosis industry. Henceforth, the medical AI pathology industry has ushered in a milestone new development.
With the rapid advancement of science and technology, pathology has long entered the era of digital pathology, and “digital pathology + artificial intelligence” has become a widely recognized trend in the industry.Digital Pathology + AIThe emergence of this technology can effectively reduce resource waste in pathological diagnosis, improve the quality and efficiency of pathological diagnosis, and alleviate the developmental challenges facing the field.
In China’s pathology testing sector, a severe shortage of pathologists (including physicians and technicians) coupled with low levels of departmental automation has resulted in an extremely heavy workload and staff overload. This situation not only hinders improvements in quality control and diagnostic accuracy but also poses potential medical safety risks, becoming a significant factor constraining the development of China’s pathology industry.
Cancerous lesions often occupy an extremely small area. In traditional pathological slide interpretation, pathologists must devote significant effort to excluding large areas of negative tissue. Artificial intelligence (AI), supported by technologies such as deep learning, can process slide images rapidly and in a standardized manner, filtering out negative samples and highlighting positive regions. This approach substantially reduces unnecessary reading time for pathologists while ensuring sensitivity, thereby assisting them in improving diagnostic efficiency and accuracy. In other words, while enhancing the training system for pathologists, AI can provide rapid and reliable diagnostic recommendations, enabling pathologists to analyze and interpret pathological slides more accurately and quickly. This empowers pathology departments to effectively improve diagnostic efficiency, quality, and objectivity, addressing challenges such as the shortage of pathologists, low manual diagnostic efficiency, and the lack of unified quality control management. The introduction of AI technology is expected to fundamentally resolve the mismatch between supply and demand, offering significant benefits to health risk management for the entire population.
To address the aforementioned challenges in the cervical cytology screening industry arising from traditional manual slide review under microscopy,91360, under the guidance of top-tier domestic and international clinical pathology experts, has undergone six years of dedicated research and developmentBy incorporating cutting-edge artificial intelligence technologies such as self-supervised learning, recurrent neural networks, and natural language processing, and leveraging the mining and processing of massive digital slides along with deep learning modeling, this system performs counting, localization, identification, analysis, and interpretation of abnormal cells in digital pathology images of cervical liquid-based cytology preparations, thereby achieving AI-assisted diagnosis for cervical cancer cytology screening.Clinical trial results demonstrated that the human-AI collaborative reading model achieved a sensitivity of 100% and a specificity of 96.29%, with a statistically significant 80.77% improvement in diagnostic efficiency. It became the first independently marketed software for assisted diagnosis in cervical cytology to receive regulatory approval globally.
Previously, in early August 2022, 91360 successfully shortlisted with this product.“AI Medical Device Innovation Task Force Unveiling Unit”, the product’s development has been recognized by the Science and Technology Department of the Ministry of Industry and Information Technology and the Medical Device Registration Division of the National Medical Products Administration.
Di Feng, Chairman of 91360, stated, “At this critical juncture when the nation emphasizes the active application of internet and artificial intelligence technologies to enhance cervical cancer prevention and control capabilities at the primary care level, 91360 has obtained China’s first Class III medical device registration certificate for pathology AI. This achievement demonstrates 91360’s commitment to digitalization and intelligent transformation within the specialized field of pathology, as well as its strength in AI-assisted diagnosis. We hope our products will alleviate the workload of pathologists, empower improvements in the efficiency and accuracy of pathological diagnosis, and contribute to addressing the shortage of highly skilled pathology professionals in primary healthcare institutions across China, thereby mitigating the insufficiency and imbalance in the distribution of pathological medical resources.”
In the future, 91360 will continue to play a leading and demonstrative role in empowering digital pathology infrastructure, driving the accelerated innovation and development of China’s pathology AI industry.