A New Correlation Network Approach for Modeling Mobility Parameters and Predicting Health Hazards
Advisor Information
Hesham Ali
Location
UNO Criss Library, Room 232
Presentation Type
Oral Presentation
Start Date
6-3-2015 1:45 PM
End Date
6-3-2015 2:00 PM
Abstract
Studying various aspects of mobility has been receiving significant attention of many research groups over the last two decades. The general relationship between mobility levels of humans and general health and clinical implications has been loosely established in several studies. However, not as much has been established on how mobility parameters can be used to develop mobility patterns to predict potential health hazards. With the recent explosion of devices that measure number of steps, number of active minutes etc. attention continued to be heavily biased in favor of data collection tools. In order to take full advantage of such devices, focus need to be placed on data integration and analysis. In this work, we focus on how to utilize the collected data in building a robust model based on correlation models to extract useful information from the raw mobility data. Correlation analysis allow us to model how mobility patterns of individuals in certain groups are associated with each other, and establish embedded associations and similarities among the mobility pattern of the group members. We introduce a correlation network approach as the basic modeling tool for representing various mobility parameters and predicting potential health problems. The proposed approach aims at identifying patterns or features associated with changes in health levels that can lead to medical intervention at the early stages of an emerging health hazard as part of a risk management plan. We illustrate the effectiveness of the proposed approach using a practical case study to link mobility with fatigue.
A New Correlation Network Approach for Modeling Mobility Parameters and Predicting Health Hazards
UNO Criss Library, Room 232
Studying various aspects of mobility has been receiving significant attention of many research groups over the last two decades. The general relationship between mobility levels of humans and general health and clinical implications has been loosely established in several studies. However, not as much has been established on how mobility parameters can be used to develop mobility patterns to predict potential health hazards. With the recent explosion of devices that measure number of steps, number of active minutes etc. attention continued to be heavily biased in favor of data collection tools. In order to take full advantage of such devices, focus need to be placed on data integration and analysis. In this work, we focus on how to utilize the collected data in building a robust model based on correlation models to extract useful information from the raw mobility data. Correlation analysis allow us to model how mobility patterns of individuals in certain groups are associated with each other, and establish embedded associations and similarities among the mobility pattern of the group members. We introduce a correlation network approach as the basic modeling tool for representing various mobility parameters and predicting potential health problems. The proposed approach aims at identifying patterns or features associated with changes in health levels that can lead to medical intervention at the early stages of an emerging health hazard as part of a risk management plan. We illustrate the effectiveness of the proposed approach using a practical case study to link mobility with fatigue.