Process Experimental Data - DU¶
- There are two general adjustments that are made to the experimentally measured joint kinematics and kinetics:
- Time synchronization, synchronize the kinematics and kinetics.
- CSU sign convention, convert descriptions of kinematics and kinetics to those used by the CSU lab.
These adjustments can be applied in any order, but for completeness, the data will first be time-synchronized, then the signs of the experimental data will be converted to the convention used by the CSU lab.
Time Synchronization¶
During a KneeHub meeting on April 2, 2019, a member of the DU lab (Donald Hume) mentioned that there is a potential issue in the synchronization between the experimental loads and displacements. This is an excerpt from that meeting’s minutes (found here: https://simtk.org/plugins/moinmoin/kneehub/2019-04-02)
… Don and Kevin discussed checking the data and noticed that load and displacement data may need to be resynchronized. Kevin offered providing a resynchronized data. Ahmet recommended keeping the data as is; list this issue as a data limitation at the wiki page and let everyone deal with it in their own way. In response to a question from Jason, Don noted that plots of load vs displacement indicate obvious trends related to this issue.…
This section describes how the load and displacement data was resynchronized. A plot of the applied anterior force and resulting anterior displacement demonstrates the time synchronization issue (Fig. 2).

Fig. 2 An example of the time synchronization issue. The x-axis is the index in the data point, and the y-axis shows the (green) applied anterior force and the (red) anterior displacement.
A semi-automatic procedure is used to find corresponding points in the experimentally measured load and kinematics curves. Different experimental data files are given for different laxity tests, and each file will be time-synchronized separately. The semi-automatic procedure plots the results from laxity test’s primary loading and kinematic direction (i.e. anterior force and anterior displacement for the anterior-posterior drawer test). The user will select a portion of the load curve that appears to be the peak load for the test (Fig. 3), and the corresponding part of the kinematics curve. A knee finding algorithm is used to determine the index of the “knee” for each curve, and the difference between these indices is used to define the kinematics curve’s offset (1),
where \(k_{load}\) is the index of the knee point in the selected portion of the load curve, and \(k_{kinematics}\) is the index of the selected portion of the kinematics curve. At least least three different offsets are defined at different parts of the curves, and the average offset is used to adjust the indices in the kinematics data (note that the average offset is rounded to an integer if needed). The offset is used to define the number of rows of kinematic data that are removed from the beginning of the .csv files.

Fig. 3 (top) The anterior force during the entire anterior-posterior drawer test, and (red) a portion of the curve that appears to be the peak load for one of the anterior drawer tests. (bottom) The selected portion of the anterior force curve, and (black) the “knee” point of the selected portion of the curve.
Time Synchronization - “Knee” Finding Algorithm¶
To avoid user variability, a “knee” finding algorithm is used to determine the knee point in a given curve. These knee points are used to find the kinematics that correspond with the applied joint loads. The algorithm that is used is from this website: https://dataplatform.cloud.ibm.com/analytics/notebooks/54d79c2a-f155-40ec-93ec-ed05b58afa39/view?access_token=6d8ec910cf2a1b3901c721fcb94638563cd646fe14400fecbb76cea6aaae2fb1
Kinematics and Kinetics Adjustment - CSU convention¶
- The joint kinematics and kinetics are converted to the convention used by CSU. This adjustment is made to keep the experimental descriptions of motion consistent the the descriptions of motion output by the knee model. This step is not strictly necessary, however this keeps the data consistent with the modeling workflow used at CSU. The CSU convention describes motions as:
- Medial tibial translation
- Anterior tibial translation
- Superior tibial translation
- Flexion
- Varus
- Internal tibial rotation.
- To convert to the CSU convention, the sign of the kinematics for three motions are changed (Table 1):
TF ML (mm)
kinematics are multiplied by -1 to convert to medial tibial translation.TF VV (deg)
kinematics are multiplied by -1 to convert to varus rotation.TF IE (deg)
kinematics are multiplied by -1 to convert to internal tibial rotation.
- To convert to the CSU convention, the sign of the kinetics for three loads are changed (Table 1):
Force TF ML (N)
kinetics are multiplied by -1 to convert to medial tibial drawer force.Torque TF VV (Nmm)
kinetics are multiplied by -1 to convert to varus torque.Torque TF IE (Nmm)
kinetics are multiplied by -1 to convert to internal tibial rotation torque.
Reported | Reported Positive Direction | CSU convention |
---|---|---|
TF ML (mm) | Lateral tibial translation | -1 |
TF AP (mm) | Anterior tibial translation | 1 |
TF SI (mm) | Superior tibial translation | 1 |
TF FE (deg) | Flexion | 1 |
TF VV (deg) | Valgus | -1 |
TF IE (deg) | External tibial rotation | -1 |
Force TF ML (N) | Lateral tibial drawer force | -1 |
Force TF AP (N) | Anterior tibial drawer force | 1 |
Force TF SI (N) | Distraction force | 1 |
Torque TF FE (Nmm) | Flexion torque | 1 |
Torque TF VV (Nmm) | Valgus torque | -1 |
Torque TF IE (Nmm) | External tibial rotation torque | -1 |
Note
There is no conversion needed for right or left knees because the kinematics and kinetics are measured and applied using the connector elements defined in the Model Development documentation. These connector elements compose the joint coordinate system ([GS83]), and their definition in developing the model takes into account whether the specimen is a right or left knee.
[GS83] | E. S. Grood and W. J. Suntay. A Joint Coordinate System for the Clinical Description of Three-Dimensional Motions: Application to the Knee. Journal of Biomechanical Engineering, 105(2):136–144, May 1983. URL: http://dx.doi.org/10.1115/1.3138397, doi:10.1115/1.3138397. |