Araştırma Makalesi

Frequency Domain Features of Acceleration Signals to Evaluate Fall Risk of Elderly

15 Ağustos 2020
PDF İndir
TR EN

Frequency Domain Features of Acceleration Signals to Evaluate Fall Risk of Elderly

Abstract

Every year, about 28-35% of people aged 65 and over in the world fall at least once, and this number will increase rapidly in the coming years. Apart from physical injury after fall, it is seen in post-fall syndromes such as addiction, loss of autonomy and depression after treatment. The impact of the fall on the individual and economically on the society is at a level that cannot be ignored and is gradually increasing. However, the fall can be prevented by correct approaches. In order to prevent fall in the elderly, balance assessment should be done frequently and precaution should be taken for individuals at risk of fall. A wide range of tools are available for balance assessment, from simple questionnaires to complex computerized tests. However, questionnaire are subjective. Computerized tests, on the other hand, are not suitable to be used in primary health care centers due to their cost and volume. Therefore, it is very important to develop a simple method that can be used in primary health care centers. Accelerometers have taken their place in wearable technology with their light, cheap and simple structures and can be used in balance assessment. In this study, it was tried to find parameters that define fall risk of the elderly by using the three axis acceleration signal recorded from the elderly with 38 non-faller 35 faller in PhysioNet database. For this, the component caused by gravity was first removed from the acceleration signal, filtered with 0.5 Hz high pass and 25 Hz low pass filter and normalized to maximum. Then, power spectrum density of acceleration signals were found using autoregressive model with 25 model order by Burg's algorithm. 29 features were obtained for all three axes, namely the descriptive features of the first and second dominant peaks in the power spectrum, the statistical features of the power spectrum, and the features related to the energy of the power spectrum. These features were compared using the independent-sample t-test at 99% confidence level. As a result, it was observed that a total of four different features showed statistically significant difference between the two groups. Among these features, the kurtosis of the power spectrum and the width of the second largest hill are added to the literature with this study.

Keywords

Kaynakça

  1. Altunkaya, S., Kara, S., Gormus, N., & Herdem, S. (2013). Comparison of first and second heart sounds after mechanical heart valve replacement. Computer Methods in Biomechanics and Biomedical Engineering, 16(4), 368-380. Retrieved from ://WOS:000317722400003
  2. Balaban, Ö., Nacır, B., Erdem, H. R., & Karagöz, A. (2009). Denge Fonksiyonunun De¤erlendirilmesi. Journal of Physical Medicine and Rehabilitation Sciences, 12, 9.
  3. Castellini, G., Gianola, S., Stucovitz, E., Tramacere, I., Banfi, G., & Moja, L. (2019). Diagnostic test accuracy of an automated device as a screening tool for fall risk assessment in community-residing elderly: A STARD compliant study. Medicine (Baltimore), 98(39), e17105. doi:10.1097/MD.0000000000017105
  4. Cho, C. Y., & Kamen, G. (1998). Detecting balance deficits in frequent fallers using clinical and quantitative evaluation tools. Journal of the American Geriatrics Society, 46(4), 426-430. Retrieved from ://WOS:000072973100004
  5. Diego Galar, U. K. (2017). eMaintenance (U. K. Diego Galar Ed.): Academic Press.
  6. Goldberger, A. L., Amaral, L. A., Glass, L., Hausdorff, J. M., Ivanov, P. C., Mark, R. G., . . . Stanley, H. E. (2000). PhysioBank, PhysioToolkit, and PhysioNet: components of a new research resource for complex physiologic signals. Circulation, 101(23), E215-220. doi:10.1161/01.cir.101.23.e215
  7. Howcroft, J., Kofman, J., & Lemaire, E. D. (2013). Review of fall risk assessment in geriatric populations using inertial sensors. Journal of Neuroengineering and Rehabilitation, 10. doi:Artn 9110.1186/1743-0003-10-91
  8. Karantonis, D. M., Narayanan, M. R., Mathie, M., Lovell, N. H., & Celler, B. G. (2006). Implementation of a real-time human movement classifier using a triaxial accelerometer for ambulatory monitoring. Ieee Transactions on Information Technology in Biomedicine, 10(1), 156-167. doi:10.1109/Titb.2005.856864

Ayrıntılar

Birincil Dil

İngilizce

Konular

Mühendislik

Bölüm

Araştırma Makalesi

Yayımlanma Tarihi

15 Ağustos 2020

Gönderilme Tarihi

28 Haziran 2020

Kabul Tarihi

10 Ağustos 2020

Yayımlandığı Sayı

Yıl 2020

Kaynak Göster

APA
Altunkaya, S. (2020). Frequency Domain Features of Acceleration Signals to Evaluate Fall Risk of Elderly. Avrupa Bilim ve Teknoloji Dergisi, 150-155. https://doi.org/10.31590/ejosat.779590

Cited By