In my UPenn graduate training, I led a multi-disciplinary team in the Jonathan Rothberg Catalyzer competition, inventing a low-cost dermatoscope prototype that was of sufficient quality to image the majority of skin lesions without a doctor visit. My key insight was that expensive optics could be replaced with modern CNN architectures, which a recent study out of Stanford validated.
I'm thankful for my diverse educational training, and in this project much of it came together: I interpreted the optics capabilities in smart phones (computer vision, UPenn), knew how to create circularly polarized light effectively and cheaply (physics, Duke/ UT Austin), fabricated prototypes iteratively and quickly (design, Harvard/ UCSC), and understood the imaging network that would deliver reliable diagnoses (machine learning, UPenn).
Skin cancer - despite being the most preventable form of cancer - still claims thousands of lives annually because of cost and access barriers. Lowering these barriers is no easy task, but ideas like simple scope may simplify the problem.
To learn more about the Rothberg Catalyzer competition at Penn, check out this video. You'll also see yours truly - we won second prize!