Information about a piece of news titled Joint moment artifacts – common errors of 3D motion analysis
Joint moment artifacts – common errors of 3D motion analysis
Intoduction
A new OSTRC study demonstrates how a common error may affect results and change conclusion of biomechanical studies of high-impact movements. Joints moments, essential to such analyses, are prone to artifacts when incorrect methods are applied in the filtering of raw data.
Joint moments are common outcome variables in studies to understand the causes of sport injuries. They are thought to represent the net effect of muscles and ligament acting across the joint, and explain the joint loading.
At the OSTRC, we use knee joint moments in our studies on ACL injury risk where athletes are tested in a state-of-the-art motion analysis lab.
Biomechanists Tron Krosshaug and his PhD student Eirik Klami Kristianslund calculate the joint moments using custom-made software.
During the course of method development, they discovered some strange-looking plots.
– We saw sharp spikes early in the joint moment curves, and could not believe that these artifactual spikes represented actual joint loading, tells Kristianslund. - After reviewing the literature, we realized that we had used the wrong methods to filter the data.
Mix of matched and unmatched cut-off frequencies found in the literature
The artifacts in joint moments arise when different (unmatched) cut-off frequencies are used for the filtering of force and movement data in the calculation of joint moments of high-impact movements. Unfortunately, the use of unmatched cut-off frequencies has been common in biomechanical studies. Kristianslund and Krosshaug worked with biomechanist Ton van den Bogert to demonstrate the effect artifacts may have on results, and published their findings in the Journal of Biomechanics.
Sidestep cutting – a typical high-impact sporting task
Recordings of sidestep cutting by 123 elite female handball players were used for this methodological study.
The hip and knee joint moments of the motion were calculated with four different combinations of cut-off frequencies for the filtering of force and movement data, representing typical methods used in previous research.
Different cut-off frequencies affected joint moments
The biomechanists identified significant differences between conditions, and as expected the greatest differences were seen between conditions with different cut-off frequencies for force data and marker data.
Hip joint moments were most affected, but substantial effects were also evident for knee abduction moments.
The differences between methods also affected the ranking of athletes based on knee abduction moments, indicating that it was not necessarily the same athletes that had the highest knee abduction moments with the various methods.
This implies that statistics of e.g. group comparisons to investigate injury risk may be different with matched or unmatched cut-offs. The authors concluded that the artifacts may have affected a wide range of biomechanical studies.
Four practical tips for clinicians interested in biomechanics
In an editorial in the British Journal of Sports Medicine, the three authors demonstrate the theoretical basis for impact artifacts and present a summary of their Journal of Biomechanics article. They also provide four practical tips for clinicians interested in biomechanics:
1. Check the filtering of markers and force data in articles you read—are the cut-off frequencies matched? The combination of high cut-off frequency or no filtering for force data and a lower cut-off frequency for marker data can potentially yield artefacts.
2. Consider if the movement studied can be considered of high impact. The impact peak present in force and segment acceleration data in high-impact movements is the basis for the described artefacts.
3. Consider if the variable of interest can be affected by artefacts. Hip joint moments are most susceptible to these artefacts, but also knee joint moments can be affected.
4. When comparing results between motion analysis studies—check if filtering methods are compatible. Do not compare magnitudes of joint moments between studies with matched and unmatched cut-off frequencies.