EECS PhD Student Rawan Alharbi Wins Best Paper Award at 2017 Obesity Weeks eHealth/mHealth Section

For the 2nd consecutive year, the HABits Lab has earned the esteemed accolade for their work on passive sensing devices for studying eating behavior.

Rawan Alharbi

For the 2nd consecutive year, the Health Aware Bits (HABits) Lab has won the Annual Paper Award for Excellence in Science at Obesity Weeks eHealth/mHealth section, held Oct. 29 - Nov. 2nd in Washington D.C. This year, EECS PhD Student Rawan Alharbi received the award.

Their paper, titled, "Will Participants Wear Passive Sensing Devices Long Enough to Study Eating Behavior?", also featured co-authors Dr. Angela F. Pfammatter (Assistant Professor of Preventive Medicine, Northwestern university) and Dr. Bonnie Spring (Chief of Behavioral Medicine in the Department of Preventive Medicine, Northwestern university), and Prof. Nabil Alshurafa (Director of the HABits Lab and Assistant Professor of Preventive Medicine and of Computer Science).

Prof. Alshurafa commented on the progress of his groups work, "Our lab is definitely breaking ground in the area of passive sensing and obesity research. This research has really improved our ability to build effective machine learning models by enabling accurate fine-grained collection of ground truth data in participants in the field. We are currently preparing for a study with participants with obesity (60 people). The ultimate goal is to complement (or provide an alternative to) existing obesity treatment methods to truly detect, understand, predict and ultimately prevent an individuals problematic eating behaviors."

Paper Background: Due to high burden and bias of self-report, passive sensing of food intake systems is rising as an alternative measure to operationalize eating habit and caloric intake behavioral constructs. Existing machine learning models designed to detect eating based on studies in-lab fail to detect eating episodes in free-living populations. In order to improve our machine learning models, a sensing suite comprising a wrist- and neck-worn sensor, and a wearable video camera (with a fish-eye lens) was designed to reliably capture eating in the field.

Paper Conclusions: Participants are willing to wear a sensing suite with a video camera if the price is right. This system opens up the potential to build effective machine learning models that can reliably passively operationalize eating-related constructs in the field. Future studies will further test the system on 60 participants with obesity in the wild.

The focus of Obesity Weeks eHealth/mHealth section is to explore the role of eHealth/mHealth in obesity-related research and treatment.