Hendriks, V. (Vincent) (2009) Validation of accelerometry based activity monitoring in the home situation of elderly and Parkinson patients. thesis, Human Movement Sciences.
![[img]](https://umcg.studenttheses.ub.rug.nl/style/images/fileicons/text.png) |
Text
Hendriks.pdf Restricted to Registered users only Download (10MB) |
Abstract
Accelerometry enables a more objective measurement of activity patterns than questionnaires. It may be used in the domestic environment, especially in elderly or other important target groups like movement restricted patients (e.g. by Parkinson's disease). An activity monitoring method (Dynaport MoveMonitor,) consisting of a single unit accelerometer and an automatic data analysis software program, categorizing data as lying, sitting, standing, walking and shuffling, has been validated in elderly and Parkinson's disease patients in-home, using video observation as the gold standard. A fixed protocol encompassed the activities standing, lying, sitting, walking and shuffling. It was extended with a 20 minute period during which various activities of daily life were performed including the aforementioned activities. The overall validity of the Dynaport system with regard to the time periods the activities were detected was 81 %. Validity did not differ significantly between elderly and Parkinson patients. In both groups the category lying was highly valid while in elderly this was also the case for walking (i.e. over 90% of the detection time). In contrast, validity with regard to the detection of sitting, standing and shuffling was less valid. Standing was often misclassified as sitting and walking, while sitting was often misclassified as standing. Splitting the data in two overall categories, i.e. active (comprising standing, shuffling and walking) and passive (comprising lying and sitting), resulted in validity values around 90 %. The validity of this single unit Dynaport system is comparable to other multi-unit accelerometers and betters that of other single unit accelerometers. Future research should focus on improving the analyzing software rather than improving the accuracy of the accelerometer apparatus.
| Item Type: | Thesis (Thesis) |
|---|---|
| Supervisor name: | Kamsma, Yvo |
| Faculty: | Medical Sciences |
| Date Deposited: | 25 Jun 2020 10:50 |
| Last Modified: | 25 Jun 2020 10:50 |
| URI: | https://umcg.studenttheses.ub.rug.nl/id/eprint/1157 |
Actions (login required)
 |
View Item |