a step towards improved knee kinematics estimates

[Daniel L Benoit]

Dear All,

The reporting of lower limb kinematics is an integral part of the research of so many, in particular research that is designed to lead to clinical intervention.

As such, I would like to compliment the efforts of Tecla Bonci, Valentina Camomilla, Raphael Dumas, Laurence Chèze and Aurelio Cappozzo for their recent paper entitled "A soft tissue artefact model driven by proximal and distal joint kinematics.", J of Biomechanics in press.

As a research community it is imperative that we recognize the limitations inherent in estimating non-sagittal plane knee joint kinematics, as evidence suggests that with skin markers our estimates may, in fact, be opposite to actual tibio-femoral motions.

This study is one more step towards gaining the ability to accurately report knee joint rotations and translations using non-invasive in vivo methods; steps which are necessary for developing evidence-based clinical intervention strategies.

Respectfully,
Daniel L Benoit
University of Ottawa

[Ton van den Bogert]

Re: a step towards improved knee kinematics estimates

Thanks, Dan, for making us aware of this paper!

One of the most interesting conclusions in the paper is that (even in running) the soft tissue artifact (STA) seems mostly related to sliding and stretching of skin, and not to wobbling caused by impact. Sliding and stretching can be explained with static models that predict STA as a function of posture, which was done in this paper.

There is some bad news as well: the models to predict STA do not seem to be generalizable across movements and across subjects.

I have to make one minor correction to the paper: it states that the running data (which I provided) was collected with video cameras, but actually this was done with 16mm cine film. I have posted a clip of one of those recordings here: http://youtu.be/nf6jkyNgkwE. The clip was converted to a rather low resolution video, but you can still get a very good impression of the soft tissue artifacts during running. The wobbling of skin markers catches the attention but may not be the most important source of error.

Dan, can you provide a reference for the evidence that bone motion can be opposite to skin marker based joint motion? We did not see that in running (Reinschmidt et al., J Biomech 30: 729-732, 1997), but I would be very interested if this has been shown for any other movements.

Now speaking as moderator of Biomch-L: I would like to see more threads like this, where recently published papers are discussed.

Ton van den Bogert

[Arnaud Barré]

Re: a step towards improved knee kinematics estimates

I could maybe make a comment regarding the bone motion and the skin marker based joint motion. We recently published a paper (Barré et al., 2013, IEEE Trans Biomed Eng, 60(11):3131-40) showing the difference of bones motion (knee prosthesis tracked by a bi-plane fluoroscopic system) and segment motion (thigh and shank tracked by a stereophotogrammetric system) during a treadmill gait task on 19 subjects.

Figure 5 shows the differences for knee joint kinematics normalized by gait cycle. Figure 6 shows the difference of the thigh (shank) motion relatively to the femoral (tibial) prosthesis motion. We can see two opposite patterns for the knee internal/external rotation angle (figure 5, row 1, column 3). If we look for the reason of this difference on the segments (figure 6), we can focus on the the vertical axis of the thigh (delta beta, row 1, column 2) which shows mostly an internal rotation of the whole segment during the first 0-10% of the gait cycle (GC), and then relax between 10-40%GC.

I hope this could help you.

Arnaud Barré

[Ton van den Bogert]

Re: a step towards improved knee kinematics estimates

Arnoud,

Thanks for this excellent reference!

Ton

[Aurelio Cappozzo]

Re: a step towards improved knee kinematics estimates

Although the potential of quantitative movement analysis in professional decision-making and intervention practice is fully recognized, its application is currently limited by several problems one of which, as opportunely highlighted by Daniel Benoit in his recent post, is the so-named soft tissue artefact. At this point in time, virtually all movement analysts are aware of the fact that this source of error severely affects their studies. However, despite the efforts of some research groups, a satisfactory solution is as yet unavailable.
Overcoming this stalemate requires the creation of new techniques and knowledge through the fusion of contributions from past and ongoing R&D programs and by fostering novel conceptual approaches for future activity. On the application front, since any departure from the status quo entails investment and generates an implicit resistance from the professions involved, the acceptance and adoption of novel procedures requires purposeful and distributed actions.
Thus an initiative that links centres of excellence in the field of human movement sciences and concordant domains is not only timely but also essential to solve the problem identified above.
The recently founded Interuniversity Centre of Bioengineering of the Human Neuromusculoskeletal System (www.iuc-bohnes.eu; under construction) would be happy to collect manifestations of interest in this direction and, should a critical mass be accomplished, promote a networking exercise.
Data sharing is also a very important issue that would help movement analysis methodologies advance.
Following Ton van der Bogert’s and colleagues’ noble example in providing us with their in-vivo data, which made our study Bonci et al., in press in the Journal of Biomechanics, possible (thanks again Ton for the data and thank you Daniel for commenting on the study), we are willing to make the ex-vivo data utilized in the latter study and in other previous publications of ours, available to the community. This data will soon be published in the above-mentioned website.
We take this opportunity for apologizing with Tom to have been imprecise in reporting the experimental set-up of his study.
Kind regards
Aurelio Cappozzo
Interuniversity Centre of Bioengineering of the Human Neuromusculoskeletal System
University of Rome “Foro Italico”

[Edward Morra]

Re: a step towards improved knee kinematics estimates

I came across this 1948 JBJS article on the shelf of our library.

TRANSVERSE ROTATION OF THE SEGMENTS OF THE LOWER EXTREMITY IN LOCOMOTION

A. S. Levens ; Verne T. Inman ; J. A. Blosser
J Bone Joint Surg Am, 1948 Oct;30(4):859-872


In which pins where "drilled firmly" into the pelvis femur and tibia of 29 subjects (ahh the good old days), and the absolute and relative rotations of the segments along their long axis was determined.

I thought it may be relevant to readers interested in this thread of discussion. Is this a well known reference or perhaps it is buried treasure?

- Ed -


Edward Morra, MSME
Manager of Computational Testing
Orthopaedic Research Laboratories, Inc.