When Susan Conover was 22, she was diagnosed with melanoma. It was at Stage II, meaning that it had grown down into the dermis layer of the skin; it likely had not yet spread, but it could potentially metastasize.
"It was basically like, 'We don't think it's gone into other parts of your body, but we don't know,'" Conover recalled.
Conover had the aberration removed. A month and a half later, she was diagnosed with melanoma again. Six months after that, she was diagnosed with melanoma for a third time. Conover has not needed cancer treatment since then, but she's had about 20 excisions of various moles in just a few years.
For Conover, it was a galvanizing experience.
"I think when you're 22, you're thinking, 'What job do I want after college' and 'who do I want to date' and 'what do I want for dinner,'" Conover said. "You're not thinking, 'Where is my life going, and is it meaningful?' I want to make sure my life is meaningful, and I also want to make sure no other patient goes through and gets as close to death as I did."
With an education in engineering and systems design from the University of Texas at Austin and MIT, Conover took an engineer's mindset to approach the problem. She has developed an AI-powered system that can identify and diagnose skin conditions, an invention that she thinks will disrupt the field of dermatology.
It is that technology that underpins Conover's startup, LuminDx, which she co-founded with Pranav Kuber at the end of 2017. Now a five-person startup based in Boston, LuminDx sells a smartphone app designed to be used by health care providers.
The startup was founded on the principle of patient-centric care. When Conover was working on her master's thesis at MIT, she interviewed different stakeholders in the world of health care innovation and found that patients ranked relatively low on the totem pole. Doctors, insurance groups and pharmaceutical companies all came first, she said.
This week, though, LuminDx received some validation from the health tech community. The startup closed an oversubscribed seed round at $2 million, led by Argon VC with participation from Flare Capital, Good Growth Capital and TBD Angels.
"We have a lot of really smart, amazing people who care about this problem, who invest in us and are off to support us," Conover said. "It does take a village. I'm excited and feeling supported by people who aren't just like, 'Oh, that's cute, you're a patient, but you really need to be building something to solve something for pharma.'"
LuminDx had been raising the round since February, bringing in the first tranche at $500,000 in May. Some of that funding went toward hiring a full-time machine learning engineer, who will bring LuminDx's AI diagnostic system up to the accuracy of a dermatologist.
Within the next 12 months, LuminDx also plans to build out its customer base, onboarding 30 primary care physicians, and increase the size of its database of photos to train the AI system. The funding is a major boost: It will allow LuminDx to license photo libraries from dermatology groups, a more streamlined approach than what the startup had previously taken.
"We were going down into 80-year-old dermatologists' basements and organizing Kodachrome images for scanning," Conover said. "It's a very trash-to-treasure sort of strategy. We've hit the 100,000 mark, which is great, but we still need more data. I think a lot of people don't realize that data is very hard to access."
LuminDx has gone through MassChallenge, as part of the summer 2018 cohort, and Techstars Boston, as part of the 2019 cohort.
The startup is actively seeking partnerships as it builds out its product and customer base.
