Figure 5 - uploaded by Halim Tannous
Content may be subject to copyright.
Source publication
Conventional musculoskeletal rehabilitation consists of therapeutic sessions, home exercise assignment, and movement execution with or without the assistance of therapists. This classical approach suffers from many limitations, due to the expert’s inability to follow the patient’s home sessions, and the patient’s lack of motivation to repeat the sa...
Contexts in source publication
Context 1
... we aim to achieve the system described in Figure 5. In an ideal scenario, the patient starts their rehabilitation program with a first clinical visit. ...
Context 2
... real-time, quaternion-based, extended Kalman filter was developed to fuse the outcomes of one Kinect visual sensor and two Shimmer IMU sensors. The overview of our developed fusion scheme is shown in Figure 50. ...
Context 3
... choosing the gradient descent algorithm as an orientation estimation filter for IMU sensors, data from the estimation of this filter and the Kinect were fused using an extended Kalman observer. Figure 55, Figure 56 and Figure 57 show the results of our real-time, quaternion-based, extended Kalman observer algorithm for fusion, for IMU sensors placed in the three proposed positions. Figure 58 presents the real-time knee flexion angle estimation using the three different techniques. ...
Context 4
... choosing the gradient descent algorithm as an orientation estimation filter for IMU sensors, data from the estimation of this filter and the Kinect were fused using an extended Kalman observer. Figure 55, Figure 56 and Figure 57 show the results of our real-time, quaternion-based, extended Kalman observer algorithm for fusion, for IMU sensors placed in the three proposed positions. Figure 58 presents the real-time knee flexion angle estimation using the three different techniques. ...
Context 5
... choosing the gradient descent algorithm as an orientation estimation filter for IMU sensors, data from the estimation of this filter and the Kinect were fused using an extended Kalman observer. Figure 55, Figure 56 and Figure 57 show the results of our real-time, quaternion-based, extended Kalman observer algorithm for fusion, for IMU sensors placed in the three proposed positions. Figure 58 presents the real-time knee flexion angle estimation using the three different techniques. ...
Context 6
... 55, Figure 56 and Figure 57 show the results of our real-time, quaternion-based, extended Kalman observer algorithm for fusion, for IMU sensors placed in the three proposed positions. Figure 58 presents the real-time knee flexion angle estimation using the three different techniques. The fusion output shows a better estimation, when ...
Context 7
... we compared our fusion filter between two IMUs, mounted in three different positions, and a Kinect camera, against the goniometer. The fusion algorithm was tested on 10 subjects and the error behaviors between Kinect, IMU and Kinect-IMU solutions seem to be stable and similar over all subjects (see Figure 55, Figure 56 and Figure 57). The fusion output shows a greater resemblance to the goniometer signal, as it almost overlaps it in p. 134 ...
Context 8
... we compared our fusion filter between two IMUs, mounted in three different positions, and a Kinect camera, against the goniometer. The fusion algorithm was tested on 10 subjects and the error behaviors between Kinect, IMU and Kinect-IMU solutions seem to be stable and similar over all subjects (see Figure 55, Figure 56 and Figure 57). The fusion output shows a greater resemblance to the goniometer signal, as it almost overlaps it in p. 134 ...
Context 9
... we compared our fusion filter between two IMUs, mounted in three different positions, and a Kinect camera, against the goniometer. The fusion algorithm was tested on 10 subjects and the error behaviors between Kinect, IMU and Kinect-IMU solutions seem to be stable and similar over all subjects (see Figure 55, Figure 56 and Figure 57). The fusion output shows a greater resemblance to the goniometer signal, as it almost overlaps it in p. 134 ...
Context 10
... these tests, the sensor is taken out of its box, and the electronic chip is modified to allow the use of a multimeter ( Figure 59). Three trials were done for each test to ensure the reproducibility. ...
Context 11
... we aim to achieve the system described in Figure 5. In an ideal scenario, the patient starts their rehabilitation program with a first clinical visit. ...
Context 12
... real-time, quaternion-based, extended Kalman filter was developed to fuse the outcomes of one Kinect visual sensor and two Shimmer IMU sensors. The overview of our developed fusion scheme is shown in Figure 50. ...
Context 13
... choosing the gradient descent algorithm as an orientation estimation filter for IMU sensors, data from the estimation of this filter and the Kinect were fused using an extended Kalman observer. Figure 55, Figure 56 and Figure 57 show the results of our real-time, quaternion-based, extended Kalman observer algorithm for fusion, for IMU sensors placed in the three proposed positions. Figure 58 presents the real-time knee flexion angle estimation using the three different techniques. ...
Context 14
... choosing the gradient descent algorithm as an orientation estimation filter for IMU sensors, data from the estimation of this filter and the Kinect were fused using an extended Kalman observer. Figure 55, Figure 56 and Figure 57 show the results of our real-time, quaternion-based, extended Kalman observer algorithm for fusion, for IMU sensors placed in the three proposed positions. Figure 58 presents the real-time knee flexion angle estimation using the three different techniques. ...
Context 15
... choosing the gradient descent algorithm as an orientation estimation filter for IMU sensors, data from the estimation of this filter and the Kinect were fused using an extended Kalman observer. Figure 55, Figure 56 and Figure 57 show the results of our real-time, quaternion-based, extended Kalman observer algorithm for fusion, for IMU sensors placed in the three proposed positions. Figure 58 presents the real-time knee flexion angle estimation using the three different techniques. ...
Context 16
... 55, Figure 56 and Figure 57 show the results of our real-time, quaternion-based, extended Kalman observer algorithm for fusion, for IMU sensors placed in the three proposed positions. Figure 58 presents the real-time knee flexion angle estimation using the three different techniques. The fusion output shows a better estimation, when ...
Context 17
... we compared our fusion filter between two IMUs, mounted in three different positions, and a Kinect camera, against the goniometer. The fusion algorithm was tested on 10 subjects and the error behaviors between Kinect, IMU and Kinect-IMU solutions seem to be stable and similar over all subjects (see Figure 55, Figure 56 and Figure 57). The fusion output shows a greater resemblance to the goniometer signal, as it almost overlaps it in p. 134 ...
Context 18
... we compared our fusion filter between two IMUs, mounted in three different positions, and a Kinect camera, against the goniometer. The fusion algorithm was tested on 10 subjects and the error behaviors between Kinect, IMU and Kinect-IMU solutions seem to be stable and similar over all subjects (see Figure 55, Figure 56 and Figure 57). The fusion output shows a greater resemblance to the goniometer signal, as it almost overlaps it in p. 134 ...
Context 19
... we compared our fusion filter between two IMUs, mounted in three different positions, and a Kinect camera, against the goniometer. The fusion algorithm was tested on 10 subjects and the error behaviors between Kinect, IMU and Kinect-IMU solutions seem to be stable and similar over all subjects (see Figure 55, Figure 56 and Figure 57). The fusion output shows a greater resemblance to the goniometer signal, as it almost overlaps it in p. 134 ...
Similar publications
Conventional musculoskeletal rehabilitation consists of therapeutic sessions, home exercise assignment, and movement execution with or without the assistance of therapists. This classical approach suffers from many limitations, due to the expert’s inability to follow the patient’s home sessions, and the patient’s lack of motivation to repeat the sa...
