Skip to main content
Fig. 3 | BMC Geriatrics

Fig. 3

From: Development and piloting of a perturbation stationary bicycle robotic system that provides unexpected lateral perturbations during bicycling (the PerStBiRo system)

Fig. 3

Example of the Kinect-system function during random-unexpected perturbation training process of an 82-year-old advanced level trainee, focusing on a reactive balance response following a 20° left unannounced tilting perturbation. Figure 3a represents the balance reactive response detection from the Microsoft Kinect point of view. At the end of the calibration stage (Fig. 3b, dashed vertical black line), the computer program and the Kinect system automatically customize the PerStBiRo system to the current trainee by calculating both: 1) the individual’s upper body sway amplitude, and 2) the trainee–stationary training bicycle (STB) zero point (i.e., vertical position). The customization process – α1 and α2 angles (presented in Fig. b – green and purple lines) are recorded separately for 90 s during the second part of the calibration stage, and at the end of this stage, the individual upper body sway amplitude and the trainee–STB zero point are calculated for both angles (α1 and α2). Then the angles that show more stability and less noisy parameters are automatically selected as the angle on which the software relies to give real-time sensorimotor feedback for an effective balance reaction by returning the STB to its vertical position. After this, the balance exercise stage begins, and the trainee is exposed to a variety of repeated random unexpected perturbations (Fig. 3b, humps in the horizontal black lines). When a 20° left perturbation is executed (the gray time-line line in Fig. 3b), the computer program checks the difference between the participant’s angle (Fig. 3b – purple line α1 or green line α2) and the training bicycle’s angle (Fig. 3b – black line) and considers the body amplitude of the participant and the trainee–STB zero point to see if there has been a significance balance reactive recovery response rather than a regular paddling movement. In this case (Fig. 3b – where the gray time-line is), the PerStBiRo system detected a reactive balance reaction (a) as the stationary training bicycle tilted only 15° (Fig. 3b – black line) out of the 20° that was programmed in the training plan

Back to article page