Link to the University of Pittsburgh Homepage
Link to the University Library System Homepage Link to the Contact Us Form

Multimodal Biometric Analysis for Monitoring of Wellness

Grigoryan, Vahan (2005) Multimodal Biometric Analysis for Monitoring of Wellness. Doctoral Dissertation, University of Pittsburgh. (Unpublished)

[img]
Preview
PDF
Primary Text

Download (1MB) | Preview

Abstract

Biometric data can provide useful information about person's overall wellness. The focus of this dissertation is wellness monitoring and diagnostics based on behavioral and physiological traits. The research comprises of three studies: passive non-intrusive biometric monitoring, active monitoring using a wearable computer, and a diagnostics of early stages of Parkinson's disease. In the first study, a biometric analysis system for collecting voice and gait data from a target individual has been constructed. A central issue in that problem is filtering of data that is collected from non-target subjects. A novel approach to gait analysis using floor vibrations has been introduced. Naive Bayes model has been used for gait analysis, and the Gaussian Mixture Model has been implemented for voice analysis. It has been shown that the designed biometric system can provide sufficiently accurate data stream for health monitoring purposes.In the second study, a universal wellness monitoring algorithm based on a binary classification model has been developed. It has been tested on the data collected with a wearable body monitor SenseWear®PRO and with the Support Vector Machines acting as an underlying binary classification model. The obtained results demonstrate that the wellness score produced by the algorithm can successfully discriminate anomalous data.The focus of the final part of this thesis is an ongoing project, which aims to develop an automated tool for diagnostics of early stages of Parkinson's disease. A spectral measure of balance impairment is introduced, and it is shown that that measure can separate the patients with Parkinson's disease from control subjects.


Share

Citation/Export:
Social Networking:
Share |

Details

Item Type: University of Pittsburgh ETD
Status: Unpublished
Creators/Authors:
CreatorsEmailPitt UsernameORCID
Grigoryan, Vahanvvgst@pitt.eduVVGST
ETD Committee:
TitleMemberEmail AddressPitt UsernameORCID
Committee ChairConstantine, Gregory Mgmc@euler.math.pitt.eduGMC
Committee CoChairChiarulli, Donald Mdon@cs.pitt.eduDON
Committee MemberErmentrout, G. Bardbard@euler.math.pitt.eduBARD
Committee MemberHauskrecht, Milosmilos@cs.pitt.eduMILOS
Committee MemberHeath, Robert Wrwheath@pitt.eduRWHEATH
Date: 31 January 2005
Date Type: Completion
Defense Date: 6 December 2004
Approval Date: 31 January 2005
Submission Date: 6 December 2004
Access Restriction: No restriction; Release the ETD for access worldwide immediately.
Institution: University of Pittsburgh
Schools and Programs: Dietrich School of Arts and Sciences > Mathematics
Degree: PhD - Doctor of Philosophy
Thesis Type: Doctoral Dissertation
Refereed: Yes
Uncontrolled Keywords: Automated Diagnostics; Biometrics; Health Monitoring; Multimodal; Wearable
Other ID: http://etd.library.pitt.edu/ETD/available/etd-12062004-153455/, etd-12062004-153455
Date Deposited: 10 Nov 2011 20:08
Last Modified: 15 Nov 2016 13:53
URI: http://d-scholarship.pitt.edu/id/eprint/10086

Metrics

Monthly Views for the past 3 years

Plum Analytics


Actions (login required)

View Item View Item