TITLE

Objectifying Subjective Medical Assessments using Smartphone Sensors

ABSTRACT

Although most of us do not have a medical degree, there are many cases when we are able to tell that someone is sick just by using our human senses (e.g., seeing someone with a pale face, touching someone with a fever). However, our senses are often not able to discern subtle issues that could have serious consequences on a person’s health. In this talk, I will highlight how smartphone sensors can be used to detect symptoms in ways that humans cannot in order to support convenient health screening and disease management. I will cover two projects in this space. The first project is PupilScreen, an app that aims to detect traumatic brain injuries based on a person’s pupillary light reflex. The second project is BiliScreen, an app that measures the degree of jaundice in a person’s eyes through a photograph; jaundice is an indicator of damage to the pancreas, making it a key symptom in conditions like pancreatic cancer or hepatitis. To conclude my talk, I will propose tools for accelerating mobile health innovation and discuss my grand vision of incorporating ubiquitous sensing into Bayesian diagnostics.

BIO

Alex Mariakakis is a 6th-year graduate student in the School of Computer Science and Engineering at the University of Washington. He is advised by Dr. Shwetak N. Patel and Dr. Jacob O. Wobbrock. As a ubiquitous computing and human-computer interaction researcher, his work identifies applications of machine learning and computer vision on data from smartphones sensors to support people’s health and public safety. His primary research examines how mobile health sensing can be used to objectively measure symptoms that are typically judged qualitatively. His work is published to top-tier venues in his field, including CHI and IMWUT. His work has also been featured by many media outlets (e.g., BBC, National Geographic, and USA Today).

Alex received his M.S. in Computer Science and Engineering at the University of Washington. He received his B.S.E. in Electrical and Computer Engineering and his B.S. in Computer Science from Duke University. Alex is a recipient of the National Science Foundation Graduation Research Fellowship, the Qualcomm Innovation Fellowship, and the Gaetano Borriello Outstanding Student Award at UbiComp 2018. More information about Alex and his work can be found on his website: https://atm15.github.io/.