five

Extracted motion features (0.2 Hz) from multisegmental, triaxial accelerometry for NIMS

收藏
DataONE2021-04-17 更新2024-06-08 收录
下载链接:
https://search.dataone.org/view/sha256:6485208d26b8682936235b0b6dfbaa18478d85a01450212e9be65f17d27dd989
下载链接
链接失效反馈
官方服务:
资源简介:
This dataset contains the motion features, extracted from non-overlapping non-overlapping 5-second windows (0.2 Hz) of filtered accelerometry, for each unique patient identifier (UPI). The format of the files are: ['features_' + UPI + '.csv']. The contents of each file include the UPI, the Recoding Index (which identifies unique timepoints in the recording, starting from 1), the hours from ICU admission of each time point, the time of day of the corresponding timepoint, the feature type (see below), and the columns containing corresponding feature values for each of the 7 sensors: Bed (control sensor), left ankle (LA), left elbow (LE), left wrist (LW), right ankle (RA), right elbow (RE), and right wrist (RW). The feature types, encoded in the 'Feature' column, are as follows: SMA - signal magnitude area, HLF_h and HLF_l - a pair of high-frequency component (HLF_h) and low-frequency component (HLF_l) time-domain medians, MFR - median frequency, FDE - frequency-domain entropy, BPW - band power between 0.3 and 3.5 Hz, and and WVL - level 2 – 6 detail coefficients of the 5th-order Daubechies wavelet transform.
创建时间:
2023-11-19
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作