Program Schedule

Using Computer Vision and Depth Sensing to Measure Healthcare Worker-Patient Contacts and Personal Protective Equipment Adherence within Hospital Rooms

Session: Poster Abstract Session: Hand Hygiene
Saturday, October 11, 2014
Room: The Pennsylvania Convention Center: IDExpo Hall BC

Determining if a healthcare worker (HCW) has practiced hand hygiene (HH) on entering or leaving a patient room is routinely performed by a variety of technologies; detecting potential HH opportunities within the room is more difficult. Our objective is to determine the feasibility of using computer vision and depth sensing to detect HH opportunities based on patient contacts, as well as determine personal protective equipment (PPE) adherence.


We used multiple Microsoft Kinects to track the 3-dimensional movement of HCWs and their hands within hospital rooms. We apply standard computer vision techniques to recognize and determine the position of fiducial markers attached to the patient’s bed determin the location of HCW hands with respect to the bed. The only data saved is in the form of Cartesian (x, y, and z) coordinates from which we can recreate skeletal paths of HCWs with respect to the bed and the room.

To measure our system’s ability to detect HCW-patient contacts we counted each time a HCW’s hands entered a virtual rectangular box aligned with a patient bed. To measure PPE adherence, we identify the hands, torso, and face of each HCW on room entry, determine the color of each body area, and compare it to the standard color of gloves, gowns and face masks, respectively. Independently, we visually examined a ground truth video recording, producing a manual count for both contact and PPE adherence. We compared our system’s results to ground truth.  


Overall, for touch detection the sensitivity was 99.7%, with a positive predictive value of 98.7%. For gowned entrances, sensitivity was 100.0% and specificity was 98.15%. For masked entrances, sensitivity was 100.0% and specificity was 98.75%, and for gloved entrances the sensitivity was 86.21% and specificity was 98.28%.


Using computer vision and depth sensing, we can estimate potential HH opportunities at the patient bedside and also estimate adherence to PPE adherence. Our approach can provide fine-grained estimates of how and how often HCWs interact directly with patients. Such approaches will help inform sub-room-level HH-monitoring efforts and other patient safety research.

Junyang Chen, BS, The University of Iowa, Iowa City, IA, James F. Cremer, PhD, Department of Computer Science, University of Iowa, Iowa City, IA, Alberto Maria Segre, PhD, University of Iowa, Iowa City, IA and Philip M. Polgreen, MD, Division of Infectious Diseases, Department of Internal Medicine, University of Iowa Carver College of Medicine, Iowa City, IA


J. Chen, None

J. F. Cremer, None

A. M. Segre, None

P. M. Polgreen, GOJO Industries: Grant Investigator, Research grant and Research support

See more of: Hand Hygiene
See more of: Poster Abstract Session
Previous Abstract | Next Abstract >>

Findings in the abstracts are embargoed until 12:01 a.m. EDT, Oct. 8th with the exception of research findings presented at the IDWeek press conferences.

Sponsoring Societies:

© 2014, All Rights Reserved.

Follow IDWeek