Used in routine vital-sign assessments, pulse oximeters determine a patient’s blood oxygen level by measuring light that travels through the finger’s skin. Studies have shown, however, that these devices can provide inaccurate readings in people with darker skin tones, leading to racial disparities in patient health.

Leveraging Penn’s physics-informed neural networks technology, Advait Thaploo C25 W25, Parthiv Patel C25 W25, and Ujjayi Pamidigantam C25 W25 are developing a way for pulse oximeter readings to account for various skin colors. Equimeter won the university’s Y-Prize competition this year as well as the Venture Lab Startup Challenge’s Jacobson Social Impact Prize.

 

Read more from the Watchlist.