Direct observation of physical activity, including posture, physical therapy, and performance of physical tasks, is the current gold standard for classifying and assessing physical activity. However, labeling, and manually assessing such observations is time consuming and expensive. Thus, there is a need for automated, cost effective tools to extract information about human physical activity from video recordings. These tools could be used for research studies as well as for remote monitoring of functional status and recovery.
In this webinar, recipients of Phase II awards for NCI's SBIR Topic 425, Information Technology Tools for Automated Analysis of Physical Activity, Performance, and Behavior from Images for Improved Cancer Health described development and validation of software tools to extract information about physical activity, posture, and aspects of physical performance related to oncology and physical trauma recovery outcomes.
Speakers
PathML – Posture Analysis Through Machine Learning
Founder of SentiMetrix, Inc.,
Program Director/Principal Investigator
PathML – Posture Analysis Through Machine Learning
Associate Professor
Department of Kinesiology and Public Health
California Polytechnic State University
PathML – Posture Analysis Through Machine Learning
Professor
Department of Computer Science and Software Engineering
California Polytechnic State University
Objective measurement of the most important predictor of oncology outcomes
Computer Scientist and Associate Director
Integrated Media System Center
University of Southern California
Objective measurement of the most important predictor of oncology outcomes
Associate Professor of Clinical Medicine
Keck School of Medicine
Section Head of Thoracic and Head/Neck Tumors USC/Norris Comprehensive Cancer Center
University of Southern California
Moderators
Program Director, Health Behaviors Research Branch
NCI Behavioral Research Program
Program Director, Basic Biobehavioral and Psychological Sciences Branch
NCI Behavioral Research Program