It is rumored that the first AI Class III medical device in China has been launched. Is it really related to artificial intelligence?

Last night, news that the Coronary Blood Flow Reserve Fraction Calculation Software Product of Koya Medical had been approved for listing by the National Medical Products Administration (NMPA) sent shockwaves through the industry.

In fact, compared to well-known star projects such as Tuishang Technology, Deepwise Medical, Shukun Technology, and Silicon-based Intelligence, Koya Medical is relatively unknown. Its obtaining the first Class III certificate for deep learning AI in China is unexpected.

Koya Medical claims this is the first Class III medical device in China to be approved using artificial intelligence technology. The approval of a Class III AI medical device is a landmark event for the medical AI industry.

In September 2017, the State Food and Drug Administration (SFDA) issued a new version of the "Medical Device Classification Catalog". According to the latest classification regulations, if diagnostic software uses algorithms to provide diagnostic suggestions, only has auxiliary diagnostic functions, and does not directly give diagnostic conclusions, it is declared as a Class II medical device. If it automatically identifies the lesion site and provides clear diagnostic prompts, it is managed as a Class III medical device. The approval of Class III certificates will become a turning point for the large-scale commercialization of medical AI, which has become a consensus in the industry.

However, from the NMPA's official statement, there are no words such as "artificial intelligence" or "AI", and the statement is vague. Is this product really related to artificial intelligence? Long-term observers of medical AI investment also have doubts. Does CT-FFR count?

Regarding this issue, iiMedia Research sought verification from executives in the same industry. He stated that this is indeed considered a Class III artificial intelligence medical device. According to the deep learning review guidelines, it is considered to be so if it contains the words "deep learning". In terms of the relationship between artificial intelligence and deep learning, deep learning is a subset of artificial intelligence; they have an inclusive relationship.

Relevant personnel from the National Medical Products Administration Center also indicated that this should be considered a Class III artificial intelligence medical device.

Specifically, this product uses its independently developed DeepVessel and DeepFFR deep learning algorithms to perform non-invasive functional analysis (CT-FFR) using coronary computed tomography angiography (CTA) images.

The prevalence of cardiovascular disease in China is continuously rising and has always been a major problem. According to the latest data released by the "2018 China Cardiovascular Disease Report", it is estimated that there are 290 million people with cardiovascular disease in China, with the death rate ranking first, accounting for more than 40% of resident disease deaths. Two out of every five deaths are due to cardiovascular disease.

It is understood that for the diagnosis of cardiovascular disease, the clinically common imaging method is to diagnose by observing the degree of stenosis of coronary vessels.However, in addition to this, evaluating the blood supply function of blood vessels is also key to evaluating the degree of myocardial ischemia.

CT-FFR is measured based on coronary computed tomography angiography (CCTA) image data and has the advantages of CCTA and fractional flow reserve (FFR).CT-FFR assesses the degree of stenosis of lesions from an anatomical perspective and can also determine the presence and severity of ischemia in stenotic coronary arteries from a hemodynamic perspective.

Koya Medical, the company behind this product, was established in early 2016, has completed hundreds of millions of yuan in financing, and has established cooperation with more than 200 hospitals in China. It is reported that Koya Medical's independently developed non-invasive CT-FFR deep pulse score product is the first Class III artificial intelligence medical imaging product to Enter the "Green Channel for Innovative Medical Devices" of the National Medical Products Administration (NMPA). Last year, products from Eagle Eye Medical, Shukun Technology, and Silicon-based Intelligence, including their retinal and coronary products, also entered the NMPA's "Green Channel for Innovative Medical Devices".

Mentioning AI, since the victory of AlphaGo in 2016, this trend has never faded. Specifically in the medical field, the medical AI industry is also exceptionally "lively", with numerous participants. Initially, doctors and hospitals were cautious in application and procurement, and compliance was one of their main concerns.With official certification from regulatory agencies such as the NMPA, medical AI products have crossed the compliance threshold and may Enter a period of market value verification.

Attached is the NMPA's official statement:

Recently, the National Medical Products Administration, after review, approved the registration of the innovative product "Coronary Blood Flow Reserve Fraction Calculation Software" produced by Beijing Kunlun Medical Cloud Technology Co., Ltd. This product is based on coronary CT angiography images and consists of an installation disc and a security dongle. Functional modules include basic image operations, vessel segmentation and reconstruction based on deep learning technology, vessel centerline extraction, and blood flow reserve fraction calculation based on deep learning technology. It has obtained one invention patent, and two invention patents are in the substantive examination stage. This product uses non-invasive technology, can reduce unnecessary coronary angiography examinations, avoid unnecessary interventional surgeries, can reduce costs, alleviate patient suffering, and can be used for early diagnosis, with significant economic and social benefits. There are no similar products registered and listed domestically, and only one similar product is listed in the United States internationally. Compared with similar products at home and abroad, the performance indicators of this product are at the international leading level. The drug supervision and management department will strengthen the post-market supervision of this product to protect the safety of patients using the device.