Infervision is planning to debut a medical imaging AI solution for the whole treatment process of lung cancer, including prediction, diagnosis, treatment and prognosis, according to Chen Kuan, Infervision’s founder and chief executive officer, disclosing it in a Chinese medical forum on July 20.
The new solution has been adopted in two Chinses tertiary hospitals, verifying the feasibility of clinical practice, said Mr. Chen. While most of Chinese AI diagnosis products providers mainly focus on the diagnosis sector, their clients - hospitals - expect for an integrated system to help doctors make decisions in the whole workflow.
In the past four years, the Beijing-based startup has employed deep learning algorithms to analyze medical images from DR, CT and MRI, developing AI diagnosis solutions for lung cancer, chest-related disease, stroke, bone disease and breast cancer. To meet the demand of doctors, the company also develops a medical research platform, which becomes the passport for Infervision to build cooperation with hospitals.
At present, Infervision is operating in 32 Chinese cities, and another seven countries in regions of North America, Asia and Europe, cooperating with over 300 hospitals worldwide. Its intelligent medical solutions complete more than 38,000 diagnosis per day, according to Mr. Chen.
The ten-year survival rate reaches 92% to date for the patients in stage 1 lung cancer - stage 1 lung cancer means that a tumor is discovered, but it is still localized within the lungs. However, only around 4.2% of the patients in stage 4 can be alive after five years since they start treatments. Thus, detecting lung cancer’s risks in the early stage is the key to the successful disease treatment, said Wang Chen, dean of Peking Union Medical College.
Chinese government issued a document for the development of the nationwide medical and healthcare careers on July 15, setting a goal that the five-year survival rates of all types of cancer should be above 43.3% in 2022 and 46.6% in 2030. There’s a long way to go obviously.
Lung cancer is the targeting disease for most of the Chinses AI diagnosis tools developers, because of the sufficient medical imaging records generated from Chinese lung cancer patients. Both the incidence rate and mortality rate of lung cancer rank the first among all types of cancers, 11.6% and 18.4% respectively in the world.
In 2017, more than 100 companies tapped a huge need for lung cancer accurate medical diagnosis, due to a large number of patients in China and a shortage of well-trained doctors able to diagnosis accurately, according to Zou Guowen, partner of Dalton Venture, a Chinses medical-oriented VC institutes.
Investors are optimistic about the potential use of AI to recognize cancer, with more than a dozen companies in China raising USD 142 million in venture capital and private equity funding in 2017 and 2018, according to KPMG.
Infervision is regarded as a dark horse in the AI medical area, attracting reputable investors to join the funding game, such as Sequoia Capital China, CDH Investments, Qiming Venture Partners. The total funding of the company in the past four years surpasses CNY 600 million (USD 87.18 million).
However, the China Food and Drug Administration (CFDA) upgraded the Medical Device Classification Catalog last August, putting AI products for diagnosis into the list of Class III devices. The companies producing such products have to apply for the Class III device licenses, which can take more than two years to obtain.
The AI diagnosis products can't be listed in the hospitals' procurement catalog without the license, resulting in obstacles of commercialization because selling products into hospitals is a mature business in the market. No company received a license to date, according to Cao Luqian, investment manager of Jiangmen, a fast-growing VC company led by the former chief executive officer of Microsoft Accelerator China.
The regulatory process is also pricey, costing up to CNY 4 million (USD 581 thousand), said Zhong Xin, chief executive officer of 12 Sigma. “It takes more money and time to get a license and many small companies will find it hard to get their products into the market,” said Mr. Zhong, “It’s not easy for newcomers.”