LABINABOX semiautomatic tracking of activity

Authors: Nadir Weibel Steven Rick Colleen Emmenegger Shazia Ashfaq Alan Calvitti Zia Agha

Publish Date: 2014/09/28

Volume: 19, Issue: 2, Pages: 317-334

PDF Link

Abstract

Patientcentered healthcare and increased efficiency are major goals of modern medicine and physician–patient interaction and communication are a cornerstone of clinical encounters The introduction of the electronic health record EHR has been a key component in shaping not only organization clinical workflow and ultimately physicians’ clinical decision making but also patient–physician communication in the medical office In order to inform the design of future EHR interfaces and assess their impact on patientcentered healthcare designers and researchers must understand the multimodal nature of the complex physician–patient–EHR system interaction However characterizing multimodal activity is difficult and expensive often requiring manual coding of hours of video data We present our LabinaBox solution that enables the capture of multimodal activity in realworld settings We focus here on the medical office where our LabinaBox system exploits a range of sensors to track computerbased activity speech interaction visual attention and body movements and automatically synchronize and segment this data The fusion of multiple sensors allows us to derive initial activity segmentation and to visualize it for further interactive analysis By empowering researchers with cuttingedge data collection tools and accelerating analysis of multimodal activity in the medical office our LabinaBox has the potential to uncover important insights and inform the next generation of Health IT systemsThis research was funded by AHRQ Grant 1 R01 HS02129001A1 Zia Agha PI We would like to thank all participants of our ongoing studies physicians and patients without which we could not collect the rich data that we used as a baseline for designing the LabinaBox Many thanks also to our colleagues participating in the QUICK and PACE research who helped the development of our tools with their advices Also thanks to Shimona Carvalho for working on an early version of ChronoSense and to Jenny Tsao for her work on expanding it A special acknowledgment to Adam Fouse the designer and developer of ChronoViz who made possible the tight integration of the many data stream collected with LabinaBox with the development of ChronoViz templates and finally to Jim Hollan Ed Hutchins and the DCogHCI lab at UCSD who were instrumental in bootstrapping this line of research

Keywords: