Subtle Medical Founder Gong Enhao: Pursuing Clinical Value to Unlock the Optimal Path for Enterprise Growth

In 2017, when the spark of artificial intelligence in the medical field was just beginning to ignite, Subtle Medical (hereinafter referred to as “Subtle Medical”) was established as an AI medical imaging company. It leveraged AI technology to enhance examination efficiency and reduce radiation dose by nearly 100-fold.

Within months of its founding, the company was accepted into Stanford’s StartX incubator and became a key project in StartX Med. The startup has garnered support and recognition from top U.S. venture capital firms, including Bessemer Venture Partners, Data Collective (DCVC), Fusion Fund, Breyer Capital, Foothill Ventures (formerly Tsingyuan Ventures), and Baidu Venture Capital.

In 2018, Subtle Medical competed against more than 3,000 companies worldwide in the “Inception AI Startup Competition” hosted by AI giant NVIDIA and emerged as the overall winner, claiming the championship in the medical applications startup category. Approximately one year after its establishment, Subtle Medical became the first China-based AI medical imaging company to have an AI imaging product receive FDA clearance and enter the U.S. market.

As the company’s technology has matured, Subtle Medical secured its Series A financing in 2020 from seasoned healthcare investors including Bencao Capital, Data Capital, and Legend Star, raising a total of approximately USD 20 million. Today, Subtle Medical has established cutting-edge collaborations and commercial implementations with hundreds of leading medical schools, hospitals, and third-party imaging centers worldwide, and is jointly developing AI and medical imaging technologies with renowned companies such as NVIDIA, Intel, GE Healthcare, Neusoft Medical, Philips China, and Siemens Healthineers.

Dr. Enhao Gong, founder of Subtle Medical, holds dual backgrounds in medicine and engineering and is a serial entrepreneur from Stanford. In him, we see a microcosm of the company itself. Continuously bridging diverse fields, Dr. Gong has undergone a personal transformation—from student to scientist, and from scientist to entrepreneur—while maintaining control over the rhythm of his own life.

Also lingering behind Gong Enhao are the questions: How can a scientist successfully navigate such an identity transition? And how can one strike a balance between scientific research and commercial implementation?such issues.

In this issue, VCBeat held an insightful dialogue with Gong Enhao, founder of Subtle Medical, aiming to provide answers to these questions.

Below are excerpts from the interview:

VCBeat:During your undergraduate years, you continuously conducted scientific research in the laboratory, went to Stanford to participate in super-resolution imaging research, competed in college student entrepreneurship competitions, co-developed the image editing software Polarr, turned down a teaching opportunity to dive into the world of entrepreneurship, and founded Subtle Medical, an AI medical imaging company... Why do you keep attempting to “cross over”?

Gong Enhao:For me, exploring various possibilities is essential. From my student days to engaging in scientific research and entrepreneurship, it has been a journey of self-discovery and self-understanding, as well as a continuous process of seeking optimization through thorough exploration. Only by constantly trying new things and maintaining a spirit of exploration can one identify their true strengths and maximize their value.

VCBeat: Most people choose to do what they are good at. Why did you choose to step out of the so-called "comfort zone"?

Gong Enhao: From a personal perspective, I hope to continually embrace challenges rather than remain within my comfort zone. However, looking back, this is also the result of “big data screening”—identifying what I excel at and where I can maximize my value through continuous exploration and experimentation. Making relatively sound life choices at critical junctures is essential.

VCBeat: Is starting a business a risky endeavor for you?

Gong Enhao: No. I am quite methodical and tend to prepare in advance. I made extensive preparations before launching the startup. At that time, the team was largely assembled, with most members hailing from Stanford University; they had a strong research background, and the technology was nearly ready.

Furthermore, by thoroughly understanding the clinical value and customer needs of these projects during their R&D or incubation phases, we have made it relatively easier to develop valuable, marketable products. In the early stages of our founding, we prioritized minimizing technical risks. Subtle Medical’s first product, SubtlePET, obtained FDA and CE certifications in 2018 and 2019, allowing us to gain a comprehensive understanding of regulatory requirements.

VCBeat: Will there be particularly difficult moments during this process?

Gong Enhao: Yes. In 2020, as we were accelerating our development and fundraising efforts, we were hit by the pandemic, which prevented us from conducting sales and promotional activities in person and disrupted our market sales rhythm. However, this also served as an impetus for us to further refine our internal platform and enhance our external-facing platform. Ultimately, we achieved a fivefold growth in business volume in 2021 and successfully deployed remote support solutions for our global customers.

Entrepreneurship inevitably involves various risks, but I believe the most critical factor is to continuously mitigate these risks. By having a comprehensive plan in place, we can ensure the company’s stable and rapid growth, while also identifying areas that require optimization through risk assessment.

“Enable the team to do it, and even do it better than you.”

VCBeat:What is the key for enterprises to overcome difficulties or sustain growth?

Gong Enhao:Team. I believe that relying solely on individual decision-making or viewing issues exclusively from one’s own perspective can often lead to insufficient information and incomplete consideration. No single person can be proficient in every aspect. Therefore, I consider teamwork to be critically important.

For example, our founding team initially handled sales ourselves, laying the groundwork for early revenue generation. However, we recognized the need for more experienced leadership to drive further growth. At the end of 2020, we recruited a Chief Commercial Officer with over 30 years of marketing and sales experience in the industry. He established a comprehensive sales framework, which became the foundation for our rapid expansion in 2021. Your role is to effectively transfer your knowledge and methodologies to the team, empowering them to execute independently—and even outperform you.

VCBeat:Do you have any experience to share on building corporate teams?

Gong Enhao:From a technical perspective, Stanford has a substantial number of researchers specializing in artificial intelligence, machine learning, and clinical technologies. Given the concentrated nature of this field, most professionals are well-acquainted with one another. As we embarked on this endeavor relatively early, establishing promising research directions and achieving notable outcomes, a portion of our team members were drawn from our own research group. This includes my senior mentor during my research period; their deep technical expertise has significantly contributed to our company’s subsequent R&D efforts. Over time, Deepwise has also attracted top scientists from various universities who specialize in the niche field of AI-powered image acquisition. The academic community frequently hosts conferences, allowing us to gain comprehensive insights into and familiarity with key players in this research circle. Consequently, we have been able to conduct targeted screening and recruitment for our team at these events.

From a product sales perspective, we place significant emphasis on individual expertise. Our current head of product sales has over 20 years of global experience in MR products at GE, and our team includes members with more than 30 years of experience at companies such as Philips and Siemens. In terms of clinical deployment, we have several senior team members with clinical technologist experience in both China and the United States. For instance, our U.S. clinical lead previously served as one of the clinical operations leaders at the Stanford Imaging Center, bringing extensive industry and clinical experience. The experience and capabilities of these team members are rare and invaluable, allowing us to place full trust in them and allocate responsibilities effectively.

Of course, team building inevitably encounters various challenges. For instance, we took some detours in the early stages of selecting personnel for our China team. To address this, in 2021 we initiated a new phase of team building and role specialization in China, recruiting more specialized professionals to take charge of market sales, technology R&D, product deployment, and other key areas. Within just a few months, we observed significant team growth and a marked increase in business performance.

VCBeat:"From the founding of Subtle Medical in 2017, to obtaining FDA clearance in 2018, and now with clinical deployments of Subtle Medical’s solutions in nearly 100 hospitals and imaging centers worldwide, what relevant experiences can you share?"

Gong Enhao:This involves two aspects: one is the approval system, and the other is the implementation system.

During the approval process, I believe it is crucial to engage in as much dialogue with the FDA as possible. First, it should be recognized that regulatory agencies have their own approval procedures and may not respond rapidly to new technologies; however, they are undoubtedly committed to delivering safer services to healthcare institutions and the public as quickly as possible. Therefore, ensuring technological safety is certainly a major priority.

But how can AI technology be implemented? It requires companies to maintain continuous and thorough communication with the FDA, truly clarifying and understanding the FDA’s regulatory reasoning. For instance, by submitting pre-submission meeting requests and being invited multiple times to present at FDA workshops, we gained early and in-depth insight into the FDA’s regulatory approach for products such as SubtleMR and SubtleGAD, which helped us prepare more effectively for our submissions. Only in this way can the FDA gain a sufficiently deep understanding of your product within its own review framework, thereby minimizing the time cost associated with obtaining approval. Thus, this is not an adversarial relationship, but rather one of mutual assistance.

In fact, the safety and efficacy that they highly prioritize are now also emphasized by the clinical community and the entire industry, representing a key area requiring in-depth research. Therefore, it is essential to fully understand the regulatory agencies’ reasoning, closely align with changes in policies or regulations, and engage in continuous collaboration and exploration to meet the standards for quality and safety set by regulatory authorities.

In terms of implementation, the healthcare system is unique due to its long sales cycle, meaning that medical sales cannot be rushed. Meanwhile, successful product adoption requires deep insights into the industry. Initially, we did not have a clearly defined direction, so we reframed the key questions as whether hospitals would be willing to adopt the product and whether it could deliver clear value to payers.

Through analysis and validation, we have found that hospitals currently demonstrate a clear willingness and ability to pay in this direction, as our product ultimately delivers value to both patients and hospitals. Consequently, AI medical imaging has transitioned from an “optional” solution to a “must-have.” We have also optimized the underlying architecture of our product to better serve physicians and patients, ensuring high quality and efficiency from the physicians’ perspective. Therefore, in terms of product design, we strive to achieve compatibility with all imaging equipment, avoid disrupting clinicians’ existing workflows, and keep data localized as much as possible to meet security requirements.

Seized the optimal timing.

VCBeat:You initially had the opportunity to take on a teaching position, but you gave it up to pursue entrepreneurship. Why?

Gong Enhao:"In fact, whether taking on teaching responsibilities or starting a business, both are excellent choices in my view. It’s just that I believed the window of opportunity for entrepreneurship was shorter and the timing more rare, with the team, timing, and other factors being quite suitable."

On the one hand, there are actually many similarities between these two roles. For instance, grant applications, patent filings, team building, and research are common to both. It was precisely because of these similarities that I felt confident in taking the next step toward entrepreneurship. On the other hand, the existing research team at the time was already strong, making it a favorable opportunity to co-found a company with partners, professors, and senior alumni. My co-founder, Professor Zaharchuk, is a clinical radiologist at Stanford and the director of relevant laboratories. He brings extensive clinical experience, a deep understanding of clinical needs, and established connections with hospital leaders. Several early team members were senior alumni from Tsinghua University’s Department of Biomedical Engineering and Stanford’s Department of Electrical Engineering, who laid a solid foundation for the company’s technological framework. Such a talented team is rare, and it further strengthened my confidence in launching the venture.

VCBeat:The First Hurdle for Scientist Entrepreneurs: Mastering the Role Transition. How to Achieve This Identity Shift?

Gong Enhao:As I mentioned earlier, there is actually commonality between these two roles, so it is not a complete transformation in the true sense. It’s just that in entrepreneurship, you need to think about and consider more things, not just writing code or doing research, but also how to transform technology into products, as well as a series of detailed issues such as product profit margins, how to truly realize clinical value, how to implement the product, how to pass regulatory reviews, how to build a team, etc., to ensure that all aspects of the company can accelerate development.

At Subtle Medical, our focus is not on pure theoretical research but rather on highly applied clinical technologies and application-oriented studies. Therefore, when your research must align closely with clinical practice, your objectives become very clear: it is not merely about building models or publishing papers, but about whether you can ultimately deliver clinical value and truly optimize workflows for patients and hospitals. This is a matter of technological value.

One key distinction in entrepreneurship is that you cannot focus solely on applying technology to a specific scenario; instead, you must ensure it performs well across all scenarios. This demands greater effort from founders, who must engage not only in research but also in technical development, dedicating more time to the latter. As the company grows, further planning will be required to explore emerging technologies, recruit team members, and address other evolving needs.

VCBeat:After commercialization, how should companies strike a balance between technological R&D and commercial realities?

Gong Enhao:First, maintaining product stability is more important than continuous expansion. Establishing a distributor network and other sales infrastructure often requires significant upfront investment. In the later stages, companies should prioritize product stability to deliver greater value to customers, thereby achieving steady, organic growth rather than pursuing expansion as the sole objective.

In fact, each patient served generates a certain amount of revenue. Therefore, I believe that a sustainable business model capable of delivering commercial value is one that optimizes technology to minimize technical costs, ensures technological stability, and accelerates both technological iteration and product deployment—rather than blindly reducing prices or pursuing continuous expansion.

VCBeat:How Would You Choose an Investment Firm?

Gong Enhao:It is ultimately a matter of mutual selection. Investors choose better projects, while we select investors who can truly add value to our business. From an enterprise perspective, I do not focus solely on capital; instead, I place greater emphasis on the value behind the funding. This includes whether the investor can provide support for corporate development and offer guidance in various areas. Some funds have deep expertise in the healthcare sector. Therefore, we prefer fund teams with clinical experience or relevant technical backgrounds, who understand industry dynamics and know how to best support companies and the sector. Additionally, they should possess a certain level of understanding of medical imaging, enabling them to continuously drive the growth of both the enterprise and the industry.

From “Optional” to “Essential”: Medical Imaging AI Reaches the Tipping Point for Greater Value Creation

VCBeat:2021 can be regarded as the inaugural year for the public listing of AI medical imaging companies. What is your perspective on this?

Gong Enhao：The public listing of AI medical imaging companies is a positive development for driving the growth of the entire industry. On one hand, they have laid the foundation for market expansion; on the other, this further validates the overall value of the market. For enterprises, it provides a wealth of valuable reference data.

VCBeat:AI-powered medical imaging has emerged as a relatively mature sector. After product approval, companies may still face challenges related to business models and technical capabilities. How is Subtle Medical finding its own rhythm for growth?

Gong Enhao:What Subtle Medical aims to do is to continuously tackle unresolved challenges in the industry. First, we persistently drive innovation and ensure continuous technological iteration, typically launching three to four new products each year. Second, after our products are adopted in clinical settings, we continually establish higher and better standards to promote the sustained development of the industry. In terms of product commercialization, we advance sales processes through enhanced communication and engagement, continuously refine our product sales system, and maintain the company’s sustainable growth.

Furthermore, from a holistic perspective, hospitals are in greater need of solutions that enhance overall imaging efficiency, rather than standalone algorithms or models. We are precisely positioned to improve overall efficiency and imaging quality for our clinical clients, encompassing areas such as clinical diagnosis and pharmaceutical workflow management. This market segment is relatively less crowded. As a distinctive player within this space, we have obtained recognition in both domestic and international markets. Therefore, we will leverage our strengths to further accelerate development in our core areas of expertise.

VCBeat:What are Subtle Medical’s specific plans for its next phase of development?

Gong Enhao:We prioritize process optimization at the data acquisition end and are therefore collaborating with upstream and downstream enterprises. I believe that only by integrating the entire workflow can we truly achieve overall efficiency gains for the product. Our partnerships with Siemens Healthineers and Bayer Healthcare, among others, are all aimed at accelerating the adoption of AI as a default component in medical workflows.

On one hand, we are conducting internal development; on the other, we are ensuring compatibility with other products. Our ultimate goal is to build an ecosystem that enables technology to effectively address overall efficiency improvement and quality assessment challenges.

Collaboration also opens up new, valuable avenues for research and development. For instance, as we deepen our deployment and partnerships within hospitals, we gain clear insights into the institutions’ emerging pain points, enabling us to design superior products that better serve our existing customers. This approach holds significant value for innovation and will ultimately expand the company’s overall customer base.

VCBeat:How to Ensure a High Success Rate for Startup Projects?

Gong Enhao:I believe we must continue to center our efforts on “delivering clinical value.” Specifically, this means continuously improving image quality to achieve greater accuracy in quantitative disease analysis, while also reducing radiation dose and accelerating scan speeds. These advancements not only enhance safety but also lower costs, and they must be effectively implemented in real-world practice. It is important to note that implementation goes beyond merely installing equipment in hospitals; it entails the integration of our technology into every aspect of hospital workflows. For example, we help customers refine their offerings, and our products become an integral, default component of their own standards.

In this process, enterprises should uphold the belief that products must truly become the default tools in clinical practice and deliver tangible value to clinical care, rather than merely undergoing trials, testing, and default operations. Meanwhile, such value should be perceptible to both physicians and patients, and quantifiable by payers.

VCBeat:Do you have any methodologies for understanding the industry?

Gong Enhao:We must understand customer needs from their perspective, while ensuring safety and efficacy, rather than merely delving deeper into technology. In my view, the specific technology employed is not critical; it does not even matter whether AI technology is used in this industry. While AI may offer technical advantages—such as faster development, better performance in reducing missed detections, and greater intelligence compared to other approaches—it should be recognized merely as an enabling tool. What enterprises need to do is leverage this technology to define their products, always with customer needs in mind.

We also took many detours along the way. Our earliest products were developed based on our own in-house research; while technologically advanced, they did not fully align with customer needs, requiring us to explore new directions. Later, we brought in more specialized professionals to our product team to lead product planning, with the ultimate goal of delivering value to hospitals.

Running a business often requires “beginning with the end in mind,” first clarifying a major goal for ten years down the road.

On this issue, Dr. Gong Enhao’s words still convey a sense of calm and rationality: “I have always believed that technology is merely a means. Regardless of the specific technology one specializes in, it is crucial to accurately identify demand—the true needs of the market. I am confident that with technological advancements, healthcare will ultimately move toward precision and personalization. No matter how long it takes, our task is to provide technologies with higher precision to support this transition.”

The Stanford Medical Future Association was co-founded by master’s students, doctoral candidates, and postdoctoral researchers at Stanford University who specialize in digital health. It aims to foster deep collaboration and exchange between scholars, entrepreneurs, and technical experts from top-tier institutions such as Stanford, Yale, Harvard, and MIT, and Chinese healthcare industry leaders, including entrepreneurs, investment professionals, policy experts, and media specialists, thereby driving innovation in research, education, and industrial applications within the healthcare sector.

Obtain a referral from at least one existing member, and send a third-person biography that fully demonstrates your long-term commitment to the digital health sector to the founder’s WeChat: leonstanford18.