Jeffrey Hausdorff, Patrick Purdon, Chung-Kang Peng
Published: Oct. 13, 2000. Version: 1.0.0
Please include the standard citation for PhysioNet:
Goldberger, A., Amaral, L., Glass, L., Hausdorff, J., Ivanov, P. C., Mark, R., ... & Stanley, H. E. (2000).
PhysioBank, PhysioToolkit, and PhysioNet: Components of a new research resource for complex physiologic signals.
Circulation [Online]. 101 (23), pp. e215–e220.
Walking stride interval time series are included for 15 subjects:
The file names indicate subject group: o for old, y for young, and pd for Parkinson’s disease. For young and old, the filename also includes the age in years.
Subjects walked continuously on level ground along an obstacle-free path. Stride intervals were measured using ultra-thin, force-sensitive resistors placed inside the shoe. The analog signal was sampled at 300 Hz using a 12-bit A/D converter with an ankle-worn microcomputer. Time between foot strikes was then automatically calculated using a validated method consistent with force-platform measurements.
Data collection:
You may download:
gait-data.tar
(150K) - full dataset as a UNIX tar archivegait-data.tar.gz
(47K) - compressed versionEach file includes two columns:
Data is also available in PhysioBank format with annotation (.str
) and header (.hea
) files,
for viewing or analysis using PhysioToolkit.
A mini-tutorial and database introduction is available on the Reylab website. For more details, please contact JM Hausdorff.