Interview: 'China's Medical AI Field is at the Dawn of a Jump' Says Deepwise's CEO
EqualOcean interviewed the co-founder and CEO of Deepwise to get a glance at the company's business prospects. The CEO is optimistic and self-confident about the near-future of Medical AI scene in China, particularly assured of his team.
On 28 June 2019, EqualOcean conducted an interview with the co-founder and CEO of Deepwise Healthcare (深睿医疗), Xin Qiao (乔昕), and found the opportunity to know more about the company at first hand.
Founded in 2017, Deepwise is a Beijing-based health-tech company that is using AI to diagnose several diseases, including, among others, lung cancer, breast cancer, stroke and bone age interpretation.
The company is devoted to optimizing the processes and improving the efficiency of the diagnosis precision.
Here are the key takeaways from the interview:
China's doctor scarcity issue is severe; AI-supported diagnostic products have a non-negligible demand.
Deepwise has landed its products in more than 300 hospitals, processing around 40,000 medical images every day, the CEO claimed.
Deepwise "has already started to generate revenue, and its revenues will jump once the company obtains the Registration Certificate for Medical Device III from CFDA," claims the CEO. The CEO has avoided providing any financial data of Deepwise's business or any concrete prediction for the certification date.
Operating from two offices in Beijing and Hangzhou, the company is currently working with around 300 colleagues and 200 freelancer doctors, who labels the raw data for the company.
The latest Series C in June will be used to expand the team and research facilities.
The core team members are from Baidu, Siemens and Peking University, which gives a precious technical and intellectual capability to the team. On the other hand, diversified cultures within the team might create a communication problem, and this stands as the biggest challenge for Deepwise to solve.
EqualOcean: How would you evaluate and position China's health-tech industry on a global scale?
Mr Qiao: China's healthcare demand&supply structure is considerably different from the West. While a general practitioner in the US could spend with each of his/her patients around 30 to 45 minutes, Chinese doctors generally spend around 5 minutes due to the high demand for healthcare and scarce sources in healthcare. These structural problems are a motivation for us to innovate for better solutions.
On the other hand, we're still in the process of catching up with Germany and the US, that are the world's health-tech leaders. Most of the Chinese hospitals are full of imported medical devices from these countries. In fact, we almost have caught them up in some fields such as Medical AI, however, for most of the areas we're still lagged behind, like pharmaceutical and drug developing industry. Yet, the government's support and vision are solid; therefore, I am confident that we will be catching up in the long term.
In Medical AI, our major advantage is the market size and favourable government policies. In China, we have several forms of incentives from public hospitals and government to land our products for the use of hospitals or scientific purposes. On top of that, we have a clear market need for AI supported diagnosis solutions due to the doctor scarcity, especially in the rural areas of China.
EqualOcean: Are there high barriers to entry for foreign medical AI companies to conquer the Chinese market?
Mr Qiao: Indeed, barriers are quite high for them. The most significant barrier is medical data accessibility and collection process. Processing, collecting and using medical data is even hard for some SOEs in China, it could be almost impossible for foreign companies.
In China, the medical data belongs to the hospital, and if we want to process that data, we need to put our servers into the hospital, as well. For most of the time, medical data cannot leave the walls of the hospital, and we have to adapt to the regulatory environment.
Medical data is technically a complex set of data, and it is strictly forbidden to go across the border unless there is a very particular international scientific collaboration. Therefore we generally see the capital inflow and financial support from overseas investors in China's medical AI field, but not the foreign companies itself.
EqualOcean: There are several other companies in the same field. What is the advantage of Deepwise in this competition?
Mr Qiao: Deepwise has a robust LDCT computational database and a diversified team from rooted institutions. Our core team members are the veterans of Pekin University, Badiu and Siemens, and we all have a strong industrial and technical experience, as well as very close relationships with the hospitals and research institutions in China.
We also have a 200 people freelancer doctor team who label hundreds of diagnosis entry every day, in a move to boost our machine learning process for Deepwise's decision-making algorithm. It is a transmission from the human brain to the machine.
Thanks to these, we have a vast hospital collaboration network and satisfactory feedbacks for our products.
EqualOcean: How many hospitals have you cooperated so far, and how much revenue has been generated out of these collaborations?
Mr Qiao: We have landed our products in more than 300 hospitals in China, in which more than 20 of them are the most-sought hospitals nationwide. Thanks to this network, our products are analyzing more than 40,000 medical images every day and are getting smarter as it processes more medical data; so we're shifting to higher gears in our already high 95% precision rate.
We have already started to generate revenue by several medtech solutions, such as cloud and other services. However, we haven't been able to monetize our AI products that we have been providing for our partners. This is because, in China, there is a strict registration process for medical devices, and Deepwise Healthcare still needs to get the Certificate for Medical Device III from CFDA (China Food and Drug Administration), which is needed to commercialize AI products.
We expect that we will obtain the Certificate in a near future and start to monetize our AI products, which will lead to a massive leap for our revenues and create a monetization paradise for the entire industry.
EqualOcean: What will be the latest Series C be used for?
Mr Qiao: We're going to expand our team and improve the research facilities. Right now, we have 80 computer scientists working with us at our R&D department, and there are 150 colleagues in Beijing, Zhongguancun office. What's more, we have 100 colleagues in Hangzhou office in which around 50 of them are sales-person and BD.
One of the most challenging parts of our operation is to convince the hospital to use our product due to the legal responsibility issues. The hospital always asks us "Who will bear the responsibility in case of a false diagnosis?" and the facts and data always back our answer. After they see that our precision rate is higher than their doctors' average, they are most of the time convinced to accept to use the product. This is a challenging communication process and handled by our 50 persons marketing team.
What are the prospects and challenges of Deepwise?
Mr Qiao: The biggest challenge for me is to manage this very diversified team. Our core members are from Baidu, Peking University and Siemens and all three of them represent different ecoles and cultures. Our colleagues from Peking University are thinking very academic and in a scientific methodology, while Baidu group is very creative and active; the rest is from traditional healthcare companies and fields, and are thinking traditionally. My biggest challenge is to clear the information between them and let them communicate effectively. To solve this problem, I am going to organize more team buildings. (Smiling)
In the long term, we might expand our business overseas; however, our focus is to operate full capacity in China. If we successfully could dominate the market in China, we will start to expand our business overseas and presumably starting from the "One Belt One Road" countries.