View Full Version : Re: Functional methods for the knee and ankle joint complex

07-23-2007, 09:56 PM
Dear all,
thank you to all to have raised and maintained a very interesting
discussion, on what seems now to be really among the most popular and
intruiguing, though complex, issues in human movement analysis. This has
been taken the interest of our research group here in Bologna for a long
time, since we were looking at the 'axode' at the knee of a patient operated
with a special prosthesis which allowed only one degree of freedom (II Symp
3DAHM - 1993, ... terrible results, by the way, just to start up with the

Fortunately, after my original interest fully dedicted to gait analysis, I
had to look also at in-vitro analyses of ankle complex motion, and more
recently to analyses of the natural and replaced knee joint, both in-vitro
and ... in the operating theatre, i.e. surgical navigation of total knee
replacement. In both conditions the direct access to skeletal motion is
guaranteed. I shall reveal that the general trend to achieve a functional
axis for the knee is shared also in this latter modern discipline, in
alternative to reference planes defined over anatomical landmarks (which are
very well identifiable as well at the knee in the operating theatre).

This experience unfortunately seems to suggest that any model based on a
single functional axis of rotation is far from all experimental evidence.
The instantaneous flexion axis at the knee changes considerably both in
orientation and position during passive flexion. The same was observed
clearly for the tibiotalar (J Biomech. 1999 32:111-8) and talocalcaneal
joints (J Biomech. 2001 34:805-9). In this latter, two distinct patterns
were observed even from the complex eversion and inversion, motion which is
known to occur in the three anatomical planes. The linear and angular
dispersion parameters of the IHA there reported should recommend care in
seeking the single flexion axis. Huge is the literature showing screw-home
plus rolling-back natural motion at the knee. Considering that all this is
worked out perhaps in the best experimental conditions, i.e. bone pins,
small calibration field, controlled smoothing/filtering techniques, it
should be recommended to be particularly careful in claiming reliable access
to functional axes from external instruments, i.e. skin placed reflective
markers. The other observation, perhaps even more concerning in this
respect, is that considerable different motion occurs in joints from
different subjects.

Therefore, in gait analysis, the interesting debate between more anatomical
or more functional based reference frames shall cope also with the troubles
implied in a reliable estimation of the axes in the latter option. Of course
single or double axis models for these two joints are admitted in in-vivo
gait analysis, but because of the multiaxal nature of the motion at these
joints, these shall be intended more as simplifying models than real
internal, i.e. skeletal, conditions. The real necessity to report joint
motion data to clinicians according to the functional axes also should be

As for the conditions under which the manouvres are performed for functional
axis estimations, perhaps I can provide my two-cent thoughts as well. Most
of what observed and reported for these two joints here above was under
'passive' conditions. Basically researchers tried to collect motion data
in-vitro in a 'virtually unloaded joints', mostly in specimens where only
articular surfaces and ligaments are left in place. Beyond the critical long
debate about the opportunity-necessity of these experiments, it is evident
that a preferred, however complex, motion occurs at these joints where only
articular surfaces and ligaments have a role, before any external force is
applied. Even in these conditions, the ligaments keep the surfaces in
contact and the surfaces keep a number of ligament fibres just tight. Of
course this critical mechanism is altered when a force is applied. If the
original aim of introducing functional axes in our gait analysis practice is
to report motion about axes representing best our specific joints, perhaps
it is exactly this the condition we want to deal with. In the best of future
possibilities, we might be able to work out functional axes under special
conditions, but what would this be under the conditions of the motor tasks
we are analysisng seems prevented. I have no competence to say what would
happen under pathological conditions, my guess is that this would be even
more complicated.

I think the discussion started from the kinematics fitting for global
optimum bone pose estimation. This is in fact a very valuable effort, though
not necessarily based on functional axes. We (a multicentre study) have
contributed to this recently (J Biomech. 2006 39:2087-95), perhaps with
complicate joint models, though for scopes not exacly associated to routine
gait analysis. Nevertheless, the more physiological are the joint models,
the more realistic would the final lower limb motion be estimated.

Sorry I have no solutions to offer, rather bad news to introduce to your
attention. But because these have not been reported yet, I thought it is
worthwhile. Hope the contribution to the debate might be valuable.

Kind regards,

************************************************** ***********************
Alberto Leardini, DPhil
Movement Analysis Laboratory
Director Prof. Sandro Giannini
Centro di Ricerca Codivilla-Putti
Istituti Ortopedici Rizzoli
Via di Barbiano 1/10, 40136 Bologna ITALY
tel: +39 051 6366522
fax: +39 051 6366561
email: leardini@ior.it

"Where is the Life we have lost in living,
Where is the wisdom we have lost in knowledge,
Where is the knowledge we have lost in information."
Thomas Stearns Eliot, Choruses from ''The Rock'' (1934)
************************************************** ***********************