Spotlight on: DeepHealth, StartHub's Startup to Watch

October 28, 2019

Each month as part of StartHub's Startups to Watch Program, we connect with the winning company to learn more about them. We had the pleasure of interviewing Bill Lotter, CTO and Co-Founder at DeepHealth. Here’s what we learned:

Tell us about your company. What is DeepHealth?

DeepHealth uses state-of-the-art machine learning to improve patient outcomes through better medical image interpretation. Our initial focus is screening mammography, where we are developing tools to help radiologists detect the presence of breast cancer earlier, more accurately, and more efficiently. Our goal is to improve breast cancer screening and make it accessible for all women.

What is your background?

I have a PhD in Biophysics from Harvard where my thesis focused on deep learning algorithms, with applications to neuroscience and medical imaging. Altogether, I have ten years of experience in machine learning, in domains ranging from algorithmic trading to sports analytics to healthcare.

How did you come up with the idea for DeepHealth?

Greg Sorensen, our CEO and a neuroradiologist with several decades of academic medicine and healthcare executive experience, came up with the idea for DeepHealth four years ago as deep learning was beginning to show tremendous promise in computer vision tasks for natural images. In an effort to apply these methods to medical imaging, he met our third co-founder, David Cox, who was my PhD advisor and is now the Director of the MIT-IBM Watson AI Lab.

We began to work together and we felt that screening mammography was the single best application -- because of its high volume combined with the difficult nature of interpretation and, most importantly, its immense medical significance. We achieved an initial boost of confidence in this direction when we participated in the Digital Mammography DREAM Challenge, an open data science competition with heavy participation, where we had the highest team submission score across all phases of the Challenge.

DeepHealth Team

How do you think your company is going to disrupt the market? What are the main differentiators between you and your competitors?

Machine learning/AI has the potential to transform many aspects of medical imaging, and healthcare in general. Given the high costs of healthcare, yet considerable presence of misdiagnosis, there are many opportunities for machine learning to help improve efficiency and diagnostic accuracy. With this potential, it is essential that machine learning applications work robustly across different settings and populations.

At DeepHealth, we are committed to developing software that works for all mammography modalities and for all women. We are validating our results at numerous medical centers, not just one, and ensuring we have testing on all ages and races. One of the most exciting disruptions is to raise the standard for underserved populations of women who currently do not have access to top quality mammography screening.

Who are your main investors?

We have a variety of funding sources, including angel investors, a large strategic partnership, and non-dilutive grant funding. Recently, we were awarded Small Business Innovation Research (SBIR) grants from both the National Institutes of Health (NIH) and the National Science Foundation (NSF), totaling $2.4M.

Why did you decide to start your business in Boston?

For biotech and healthcare, there is no better place to be than Boston. The world-class hospitals and schools provide a great scientific culture, ample opportunities for collaboration, and an awesome talent pool.

What resources in the startup community have contributed to the success of your company?

We were fortunate to be a member of the MassChallenge HealthTech 2018 cohort. Through the community, access to strategic partners, and various perks such as in-kind deals and office space, MassChallenge was instrumental in our early development. Additionally, we have had our offices in the Z-Park Innovation Center in Cambridge for the past year. The facilities, atmosphere, and location have been great for us.

How do you use StartHub professionally?

StartHub is a great place to find news and monitor trends in the Boston startup community, as well as discover events for networking and education.

What are the next steps for your company?

A key next milestone for DeepHealth is receiving FDA approval for our mammography software, which will allow us to launch our first product next year.

What is the most important thing our readers should know about DeepHealth?

Although there is certainly a lot of hype around AI right now, DeepHealth is committed to doing the most rigorous science to create software that clinicians and patients can trust. We firmly believe that trust is what will make us successful in the long run.

Read Complete Article