HCII Ph.D. Students Receive CMLH Fellowship in Digital Health

July 19, 2017

Four Human-Computer Interaction Institute Ph.D. students are recipients of the Center for Machine Learning and Health (CMLH) Fellowship in Digital Health. Incoming students Prerna Chikersal and Megan Hofmann join current Ph.D. students Qian Yang and Julian Ramos as winners of this fellowship.

The CMLH Fellowship is for graduate students, both Ph.D. and master's, who are pursuing cutting-edge research that supports digital health. While the fellowship leaves this broadly defined, the research projects should have, "a high potential for impact, value and innovation to solve problems encountered by various groups of people involved in health care." The HCII students selected bring research ranging from machine learning to 3D modeling, but all focused on advancing medical care.

Prerna Chikersal's project aims to develop a cyber-physical system that senses the employees' environment, work-life balance and physical and mental health to promote healthy behaviors. Her novel project combines environmental and personal sensor data, which she believes will improve employee health predictions. "Beyond the moral responsibility to send workers home healthy and whole, employers also have compelling business reasons to promote the health and wellbeing of employees," she writes in her proposal.

While measuring sleep will be one aspect of Chikersal's project, Julian Ramos' project looks even more extensively at sleep, it's impact on well being, and how to provide personalized, context-aware and effective sleep hygiene recommendations. His project means to accomplish this through a smartphone application. Persona, Ramos' application, is a system that automatically generates personalized and context-aware sleep recommendations using sensors in the patient's smartphone and a fit bit activity tracker.

"Sleep hygiene recommendations are general and do not take into account that the patient may not have the skills, preferences or ability for modifying certain habits," he writes. As an example, Ramos uses the sleep hygiene recommendation against exercising before sleep. For people who do not have the option to avoid evening exercise, he believes a better sleep hygiene tip would be more effective for this person. It is Ramos' hope that the methods developed in this project can then be applied to health topics like weight loss or stress management.

Both Ramos' and Chikersal's projects use sensing to collect data that drives health predictions or recommendations. Qian Yang hopes to use machine learning to create her set of recommendations. Rather than directing them to the patient, her project is designed to help inform health care providers for heart disease patients.

Yang and team have developed a decision support tool for healthcare providers that is radically different from other similar tools. Using machine learning, her tool mines patient information from electronic medical records and automatically combines information for cardiologists and surgeons who will decide which patients are candidates for ventricular assist devices, a treatment for late-stage heart failure. She will have the opportunity to test and evolve her design with clinicians thanks to the CMLH Fellowship.

Megan Hofmann's project is a 3D modeling tool to develop customized assistive technology for people with limb motor impairments of which there are over 700,000 people in the US alone. Hofmann has a strong research background in assistive technology including 3D printed prosthetics that can help patients who have been unable to find or afford limb replacement options. Using rapid prototyping technology, Hofmann and team can help patients create low-cost and effective prosthetics. However, during this research Hofmann realized that the current modeling tools don't fully support the creation or customization of assistive technology. "I propose to develop a 3D modeling tool that supports modular design by imposing an object-oriented design paradigm onto a widely used 3D modeling tool, Autodesk Fusion 360," she writes.

All four students receive 12 months of stipend support, a full academic year of tuition and $3,000 in funds for research expenses. Students will present their work at a CMLH-sponsored event.