Hot Searches: Color Doppler Ultrasound Probe Puncture Rack
scroll down
Latest Products
The State Food and Drug Administration is collecting information on AI medical devices; no relevant products have been registered.
Release time:
2025-03-12 09:43
Medical AI is so popular that regulations are gradually catching up.
On November 19, the Center for Medical Device Evaluation of the National Medical Products Administration (NMPA) issued a "Notice on Soliciting Information from Production Enterprises of Artificial Intelligence Medical Devices".
The notice stated that in recent years, artificial intelligence technology has developed rapidly, and products combining it with medical devices have begun to emerge. Therefore, the review center has begun to conduct relevant research.
The "artificial intelligence medical devices" mentioned in the notice specifically refer to medical devices that use "new generation artificial intelligence technology" in aspects such as product workflow optimization, data processing, and auxiliary diagnosis. "New generation artificial intelligence technology" refers to technologies, such as deep learning and neural networks, that use data-driven methods to train algorithms.
Production enterprises that have developed, produced, or applied for registration of such products are encouraged to actively participate and cooperate. As there are currently no related products applying for registration, in order to facilitate contact and further improve relevant work, relevant production enterprise information is being collected from both domestic and foreign sources.
The following companies are not included in the scope of this information solicitation: products that do not use the above-mentioned new generation artificial intelligence technologies; those that use new generation artificial intelligence technology for product research and development or lifecycle management, but not for the product itself; and those that do not belong to medical devices.
Due to the rapid iteration of AI technology, there is a huge gap with regulation. In April this year, the U.S. FDA approved IDx-DR, the first medical device that uses artificial intelligence to detect diabetic retinopathy; on May 24, Imagen's OsteoDetect software was also approved by the FDA. This software uses machine learning technology to analyze two-dimensional X-ray images and identify whether a patient has a fracture by identifying the patient's wrist anterior-posterior and lateral X-ray images.
From the perspective of domestic regulation, the path for reviewing and approving a medical AI product is still unclear, and no product has been officially approved.
Previously, according to the revised "Medical Device Classification Catalog" (hereinafter referred to as the "New Classification Catalog") issued by the State Food and Drug Administration in September 2017 and implemented on August 1, 2018, the definition of diagnostic function software appeared, which means that medical AI imaging companies have a requirement for "licensing", and can no longer remain in the hospital's "free trial" stage, putting forward new requirements for promoting the standardized development of the industry.
According to the New Classification Catalog, the risk level of diagnostic function software is determined based on the risk level, maturity level, and publicity level of the algorithm used, not only on the basis of the object being processed (such as images of cancers, malignant tumors, etc.).
If diagnostic software, through its algorithm, provides diagnostic suggestions and only has auxiliary diagnostic functions without directly giving a diagnostic conclusion, the relevant products in the diagnostic catalog are managed as Class II medical devices. If diagnostic software automatically identifies the lesion site and provides clear diagnostic prompts through its algorithm, then its risk level is relatively high, and the relevant products in the diagnostic catalog are managed as Class III medical devices.
"Domestic AI development in medical applications is much faster, but currently there are indeed no products that have passed CFDA regulatory approval. There is also a working group within the CFDA that is considering how to regulate and approve them, and there are no relevant regulations yet." Zhang Mingdong, Chief Medical Officer and Vice President of Regulatory Affairs of Boston Scientific Greater China, and former medical officer of the U.S. Food and Drug Administration (FDA) Center for Devices, previously told the 21st Century Business Herald, "China is also learning how the U.S. regulates and considering the actual situation in China. Due to differences in the speed of development, the differences in approval requirements between China and the U.S. for traditional medical devices are relatively large, but for artificial intelligence devices, everyone started around the same time, so the regulatory thinking should be very close. There are also medical device regulatory alliances internationally that are coordinating. I believe that the regulatory differences between China and the U.S. in this area will not be too large, and the synergy will be higher."
Taking the FDA's experience as an example, following the approval of the AI product IDx-DR for detecting diabetic retinopathy, the FDA also approved an imaging AI product, the OsteoDetect software. Imagen submitted a study to the FDA including 1000 X-ray images, evaluating the independent performance of OsteoDetect's wrist fracture detection image analysis algorithm and its accuracy in identifying fractures, and comparing the results of the algorithm to the judgments of three professional orthopedic surgeons. A retrospective study of 200 patients was also submitted.
"The difficulties in regulation and approval are related to the characteristics of artificial intelligence." Zhang Mingdong believes that the difficulty lies in the fact that, unlike traditional devices, "medical devices themselves are updated quickly, and AI devices are updated even faster. Therefore, the regulation of such medical devices cannot continue the traditional regulatory thinking and needs to be adjusted. However, the same principle is based on risk, that is, regardless of the type of medical device, the risk must be considered."
FDA Commissioner Scott Gottlieb, at the 2018 Health Datapalooza conference in Washington, D.C., stated that the FDA is expanding opportunities for digital health tools and actively developing new regulatory frameworks to review artificial intelligence using new methods, while protecting patients.
The FDA expects that more and more AI-based tools will be submitted for review in the coming years, with medical imaging equipment at the forefront. The FDA's approach to AI will focus on its methods of handling real-world data, including structured and unstructured data from pathology slides, electronic health records, wearable devices, and insurance claims data.
To this end, the FDA has developed and adopted the Pre-Cert plan, which allows companies to make minor changes to their devices without having to submit an application for review each time. Moreover, the FDA will ensure that other aspects of the regulatory framework (such as new software validation tools) are sufficiently flexible to keep pace with the unique attributes of this rapidly evolving field. The FDA's regulation of AI ensures that these new technologies meet their safety and effectiveness standards. Medical AI is so popular that regulations are gradually catching up.
On November 19, the Center for Medical Device Evaluation of the National Medical Products Administration (NMPA) issued a "Notice on Soliciting Information from Production Enterprises of Artificial Intelligence Medical Devices".
The notice stated that in recent years, artificial intelligence technology has developed rapidly, and products combining it with medical devices have begun to emerge. Therefore, the review center has begun to conduct relevant research.
The "artificial intelligence medical devices" mentioned in the notice specifically refer to medical devices that use "new generation artificial intelligence technology" in aspects such as product workflow optimization, data processing, and auxiliary diagnosis. "New generation artificial intelligence technology" refers to technologies, such as deep learning and neural networks, that use data-driven methods to train algorithms.
Production enterprises that have developed, produced, or applied for registration of such products are encouraged to actively participate and cooperate. As there are currently no related products applying for registration, in order to facilitate contact and further improve relevant work, relevant production enterprise information is being collected from both domestic and foreign sources.
The following companies are not included in the scope of this information solicitation: products that do not use the above-mentioned new generation artificial intelligence technologies; those that use new generation artificial intelligence technology for product research and development or lifecycle management, but not for the product itself; and those that do not belong to medical devices.
Due to the rapid iteration of AI technology, there is a huge gap with regulation. In April this year, the U.S. FDA approved IDx-DR, the first medical device that uses artificial intelligence to detect diabetic retinopathy; on May 24, Imagen's OsteoDetect software was also approved by the FDA. This software uses machine learning technology to analyze two-dimensional X-ray images and identify whether a patient has a fracture by identifying the patient's wrist anterior-posterior and lateral X-ray images.
From the perspective of domestic regulation, the path for reviewing and approving a medical AI product is still unclear, and no product has been officially approved.
Previously, according to the revised "Medical Device Classification Catalog" (hereinafter referred to as the "New Classification Catalog") issued by the State Food and Drug Administration in September 2017 and implemented on August 1, 2018, the definition of diagnostic function software appeared, which means that medical AI imaging companies have a requirement for "licensing", and can no longer remain in the hospital's "free trial" stage, putting forward new requirements for promoting the standardized development of the industry.
According to the New Classification Catalog, the risk level of diagnostic function software is determined based on the risk level, maturity level, and publicity level of the algorithm used, not only on the basis of the object being processed (such as images of cancers, malignant tumors, etc.).
If diagnostic software, through its algorithm, provides diagnostic suggestions and only has auxiliary diagnostic functions without directly giving a diagnostic conclusion, the relevant products in the diagnostic catalog are managed as Class II medical devices. If diagnostic software automatically identifies the lesion site and provides clear diagnostic prompts through its algorithm, then its risk level is relatively high, and the relevant products in the diagnostic catalog are managed as Class III medical devices.
"Domestic AI development in medical applications is much faster, but currently there are indeed no products that have passed CFDA regulatory approval. There is also a working group within the CFDA that is considering how to regulate and approve them, and there are no relevant regulations yet." Zhang Mingdong, Chief Medical Officer and Vice President of Regulatory Affairs of Boston Scientific Greater China, and former medical officer of the U.S. Food and Drug Administration (FDA) Center for Devices, previously told the 21st Century Business Herald, "China is also learning how the U.S. regulates and considering the actual situation in China. Due to differences in the speed of development, the differences in approval requirements between China and the U.S. for traditional medical devices are relatively large, but for artificial intelligence devices, everyone started around the same time, so the regulatory thinking should be very close. There are also medical device regulatory alliances internationally that are coordinating. I believe that the regulatory differences between China and the U.S. in this area will not be too large, and the synergy will be higher."
Taking the FDA's experience as an example, following the approval of the AI product IDx-DR for detecting diabetic retinopathy, the FDA also approved an imaging AI product, the OsteoDetect software. Imagen submitted a study to the FDA including 1000 X-ray images, evaluating the independent performance of OsteoDetect's wrist fracture detection image analysis algorithm and its accuracy in identifying fractures, and comparing the results of the algorithm to the judgments of three professional orthopedic surgeons. A retrospective study of 200 patients was also submitted.
"The difficulties in regulation and approval are related to the characteristics of artificial intelligence." Zhang Mingdong believes that the difficulty lies in the fact that, unlike traditional devices, "medical devices themselves are updated quickly, and AI devices are updated even faster. Therefore, the regulation of such medical devices cannot continue the traditional regulatory thinking and needs to be adjusted. However, the same principle is based on risk, that is, regardless of the type of medical device, the risk must be considered."
FDA Commissioner Scott Gottlieb, at the 2018 Health Datapalooza conference in Washington, D.C., stated that the FDA is expanding opportunities for digital health tools and actively developing new regulatory frameworks to review artificial intelligence using new methods, while protecting patients.
The FDA expects that more and more AI-based tools will be submitted for review in the coming years, with medical imaging equipment at the forefront. The FDA's approach to AI will focus on its methods of handling real-world data, including structured and unstructured data from pathology slides, electronic health records, wearable devices, and insurance claims data.
To this end, the FDA has developed and adopted the Pre-Cert plan, which allows companies to make minor changes to their devices without having to submit an application for review each time. Moreover, the FDA will ensure that other aspects of the regulatory framework (such as new software validation tools) are sufficiently flexible to keep pace with the unique attributes of this rapidly evolving field. The FDA's regulation of AI ensures that these new technologies meet their safety and effectiveness standards.
Contact Us
Room 618, 36 Gaoke Road, Tianhe District, Guangzhou, China
national unified free service hotline:400-020-6788
Copyright©2025 All Rights Reserved of Rencheng (Guangzhou) Medical Equipment Maintenance Co., Ltd.
