Linguistic Summarization of Sensor Data for Early Illness Recognition in Eldercare

University Of Missouri-Columbia, Columbia MO

Investigators

Linked publications, trials & patents

Abstract

Project Summary/Abstract Older adults and others with chronic health conditions benefit most from early detection of health changes and early treatment to prevent hospitalization or nursing home placement. Since 2005, our interdisciplinary research team has rigorously tested a state of the art sensor monitoring system for older adults in TigerPlace, a unique aging-in-place facility near the University of Missouri (MU) campus. We have shown that unobtrusive, continuous monitoring of individuals with in-home sensors provides useful embedded health assessment. While in-home sensors hold enormous potential for identifying early changes in health, analyzing the resulting data is challenging for clinicians due to its variety, velocity and volume. We propose to develop a new knowledge generation framework based on linguistic health summaries as a tool to integrate and summarize vast amounts of in-home sensor data. We investigate the innovative use of linguistic summaries for effective communication, and as a natural way to address the big data issues that arise from large amounts of health related evidence coming from heterogeneous sensors. We will apply our framework for improving early illness recognition in nursing homes. We will evaluate our knowledge generation framework in TigerPlace and 14 other Americare Inc., Sikeston, Missouri nursing homes in Missouri where we have deployed sensor networks in the last two years.