Use of wearable ambulatory monitor in the classification of movement states in Parkinson's disease

For Parkinson's patients to function at their best, their medications need to be optimally adjusted to the diurnal variation of symptoms. For this to occur, it is important for the managing clinician to have an accurate picture of how the patient's bradykinesia/hypokinesia and dyskinesia fluctuate throughout the normal daily activities. This thesis proposes the use of wearable accelerometers coupled with machine learning and statistical techniques in order to classify the movement states of Parkinson's patients and to provide a timeline of how the patients fluctuate throughout the day. A pilot study was performed using 2 patients with the goal of assessing the ability to classify dyskinesia and bradykinesia/hypokinesia based on accelerometric data. The patients were observed and videotaped. Clinical observations of bradykinesia/hypokinesia and dyskinesia were noted every minute. Neural networks were able to classify better than classification trees with an average c-index (equivalent to the area under the ROC curve) of 0.905 for bradykinesia/hypokinesia and 0.926 for dyskinesia. A separate group of 5 patients were observed with the additional goal of building models that can classify the movement of a patient without requiring clinically annotated training data for the same patient. An enhanced protocol was used in the final study. Dichotomized linear regression was found to classify well with an average c-index of 0.8219 for body bradykinesia/hypokinesia and 0.8799 using as the gold-standard the patient's diary. Dyskinesia was classified at a c-index of 0.7522. Neural networks did not perform as well, possibly because of restrictions placed on adjusting parameters. The two most clinically important problems: predicting

Department dc:contributor.department

Harvard University--MIT Division of Health Sciences and Technology.

Grantor dc:publisher

Massachusetts Institute of Technology

Year dc:date.issued

2003

Author dc:creator

• Klapper, David A. (David Asher), 1966-

Advisor dc:contributor.advisor

• Lucila Ohno-Machado.

• Harvard University--MIT Division of Health Sciences and Technology.

Statement dc:rights

• M.I.T. theses are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. See provided URL for inquiries about permission.

Licence dc:rights.uri

• http://dspace.mit.edu/handle/1721.1/7582

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/1721.1/98336

OAI identifier oai:identifier

oai:dspace.mit.edu:1721.1/98336