Use of a Single Wireless IMU for the Segmentation and Automatic Analysis of Activities Performed in the 3-m Timed Up & Go Test

Figures

• 

  Figure 1. Activities performed in the 3-m TU & Go test. 1 = standing, 2 = first walk, 3 = 3-m turning, 4 = second walk, 5 = pre-sitting turning, and 6 = sitting.

• 

  Figure 2. IMU sensor develop in the biomedical engineering laboratory of the Universidad de Concepción, Chile.

• 

  Figure 3. Set up used for the acquisition data during 3-m TU & Go test.

• 

  Figure 4. Identification of Standing/Sitting events using Pitch.

• 

  Figure 5. Peaks identified for the search of start and end of standing and sitting actions.

• 

  Figure 6. Result of the proposed method for the search of the beginning (standingi) and the end (standingf) of the Standing event and for the search of the beginning (sittingi) and the end (sittingf) of the Sitting event.

• 

  Figure 7. Automatic feature-based segmentation algorithm for the standing and sitting activities in the 3–m TU & Go test.

• 

  Figure 8. Turning events identification using Yaw.

• 

  Figure 9. Maximum and minimum of the rotation signal derived to start the search for turning events.

• 

  Figure 10. Result of the proposed method for the search of the beginning (turn1i) and the end (turn1f) of the 3-m turning event and for the search of the beginning (turn2i) and the end (turn2f) of the pre-sitting event.

• 

  Figure 11. Automatic feature-based segmentation algorithm for the turning actions in the 3-m TU & Go test.

• 

  Figure 12. Results from the measurements obtained by the proposed methodology versus the typical visual clinical procedure in young subjects. (a) Pearson correlation between IMU and the video record analysis; (b) Bland–Altman plot between IMU and the video record analysis.