HCESC Jump ARCHES Summer Interns: Zijun Yu

Twelve undergraduate interns are joining the 2022 Jump ARCHES Summer Internship Program, led by HCESC. The internship pairs students up with project teams led by UIUC faculty and OSF Healthcare/UICOMP researchers that have active Jump ARCHES awards. We've asked each intern to tell us a little about themselves. We'd like to introduce you to Zijun Yu!

Zijun is working with HDFS Professor Nancy McElwain on the Jump ARCHES project Early Detection of Developmental Disorders via a Remote Sensing Platform. Among U.S. children between 2 and 8 years of age, approximately 15-17% are estimated to have at least one diagnosed mental, behavioral or developmental disorder (MBDD), yet the ability to evaluate and treat MBDDs in the first years of life is limited. The central objective of this Jump ARCHES project is to develop novel technologies to better enable early detection and intervention of MBDDs among young children. To this end, researchers are collecting data from 120 children between 24 and 48 months of age (60 children with clinical referral; 60 children from the community) using a wearable sensing platform, LittleBeats, developed by their interdisciplinary team that has expertise in machine learning and signal processing (Mark Hasegawa-Johnson, Romit Roy Choudhury, Bashima Islam, Jialu Li), parent-child interaction dynamics and early socioemotional development (Nancy McElwain, Yannan Hu, Kexin Hu), and childhood delays and disabilities (Siraj Siddiqi, physician, UICOMP). LittleBeats integrates multiple sensors into a compact form factor and simultaneously captures speech (via audio), motion (via inertial motion sensor [IMU]), and physiological signals (via ECG sensor) of young children in their natural environments for extended periods of time and without researchers or clinicians present. They apply machine-learning approaches to data collected via LittleBeats to obtain multimodal profiles of child functioning that then can be used to detect delays and specify targets for intervention. Critical next steps in our project will be validation of motion algorithms (applied to the IMU data) and visualization of the complex multi-modal data.