PDA

View Full Version : SUMMARY - knee model advice



Carol F. Kallfelz
12-06-1999, 02:57 AM
Please find below my original posting and responses. Thanks to those
who replied.

--

Carol F. Kallfelz, MS, Engr
Supervisor, Motion Analysis Laboratory
RR&D Center for Limb Loss Prevention and Prosthetics Engineering
VA Puget Sound Health Care System
1660 S. Columbian Way MS 151
Seattle, WA 98121
Phone: (206) 762-1010, ext. 63223
Fax: (206) 764-2127
Email: kallfelz@u.washington.edu



ORIGINAL POSTING
__________________________________________________ ____________________

Dear Subscribers,

Please bear with yet another request for advice on knee models: I would

appreciate hearing from those of you who have implemented one of the
many published models of the knee to solve the indeterminate problem.

My goal is a familiar one: to predict the muscle, ligament, and
articular contact forces throughout the gait cycle using kinematic,
ground reaction force, and anthropometric data. Id prefer to use an
existing dynamic, rigid body model. Accuracy, particularly of the
contact forces, is my primary concern; ease of implementation is
secondary. Ive identified the following models/approaches as possible
candidates:

Schipplein and Andriacchi (1991) Interaction between active and passive
knee stabilizers during level walking. J Orthop Res 9:113-119

Tumer and Engin (1993) Three-body segment dynamic model of the human
knee. Transactions of the ASME 115:350-6 (modified for the inverse
dynamic problem)

Collins JJ (1995) The redundant nature of locomotor optimization laws. J

Biomech 28(3):251-267. (later modified and validated by Lu et al
(1997))

Glitsch and Baumann (1997) The three-dimensional determination of
internal loads in the lower extremity. J Biomech 11/12:1123-1131


Before committing to one, I would appreciate advice from anyone who has
had practical experience (good or bad) with any of the above. Are there

well-established benefits/limitations of each? How sensitive is the
model to anatomical detail (i.e., can generic anatomy be used with
reasonable confidence if CT/MRI of individual subjects is not possible)?

And is there a more appropriate model that I have overlooked? General
comments (of the if I knew then what I know now kind) concerning 2D
vs. 3D, phenomenological vs. anatomical models, the importance of
including the patellofemoral joint, optimization criteria, role of EMG,
etc. would also be most welcome.

Thanks in advance for your help.





RESPONSES
__________________________________________________ ____________________

There are many other models for estimating knee forces than you have
listed. For example,

Kaufman et al. (1991) Physiological prediction of muscle forces - I.
Theoretical Formulation. Neuroscience 40: 781-792.

Kaufman et al. (1991) Physiological prediction of muscle forces - II.
Application to isokinetic exercise. Neuroscience 40: 793-804

Mikosz et al. (1988) Model analysis of factors influencing the
prediction
of muscle forces at the knee. JOR 6: 205-214.

Also see older papers by Crowinshield and Brand (1981), D.E. Hardt
(1978),
Seireg and Arvikar (1973, 1975), and A.G. Patriarco et al. (1981).

For sensitivity analysis to input parameters, see R. A. Brand et al.
(1986)
The sensitivity of muscle force predictions to changes in physiologic
cross-sectional area. J Biomech 19: 589-596.

You have chosen an area with an extensive literature. It is worthwhile
doing an exhaustive literature search before deciding on a model. I
would
note that one of the main differences in modeling approach is whether
optization or EMG (or both) are used to estimate muscle forces. Many
researchers have strong feelings about which approach is superior.

Regards,

Richard Hughes, Ph.D.
Departments of Surgery and Biomedical Engineering
University of Michigan



__________________________________________________ ____________________


Carol,

Here's another model paper that might help. If you have any questions
on
this one, I would be happy to answer them:
Zheng N, Fleisig GS, Escamilla RF, Barrentine
SW.
An analytical model of the knee for estimation of internal forces during

exercises. Journal of Biomechanics 31(10):963-967, 1998.

- Glenn

A SSSS M M IIIIIIII
A A S S MM MM I
A A SS M M M I
AAAAA SS M M I
A A S S M M I
A A SSSS M M IIIIIIII
************************************************** *******
Glenn S. Fleisig, Ph.D.
work phone: 205-918-2139
work fax: 205-918-0800
address: American Sports Medicine Institute
1313 13th Street South
Birmingham, Alabama 35205
email: glennf@asmi.org
web site: http://www.asmi.org
************************************************** ********



__________________________________________________ ____________________

Dear Carol:

Part of my research in knee mechanics involved biomechanical models for
measuring joint reaction forces due to both external and internal (ie,
muscle, ligament,
tibiofemoral contact) loads. The following models are a just a few that
are in existent today and are only appropriate if your objectives are
similar to those of the
investigators who incorporated these models:

(1) Morrison, JB. (1968) Bioengineering analysis of force actions
transmitted by the knee joint. Biomedical Engineering 3: 164-170
COMMENTS: Good kinetic model but doesn't account for osteokinematic
details (ie, changes in tibiofemoral contact points)

(2) Schipplein, OD, and Andriacchi, TP. (1991) Interaction between
active and passive
knee stabilizers during level walking. J Ortho Res 9:113-119
COMMENTS: Similar model to Morrison (1968) with two unique traits: (a)
soft tissue tensions were considered in equilibrium equations; and (b)
varus/valgus
loads were balanced by collateral ligament tensions; Good model for
calculating total rxn force, especially in determining the forces
involved in condylar lift-off (ie,
increased medial compartment loading). I think there were not enough
details described regarding the derivation of equilibrium equations.

(3) Lutz, G.E., Palmitier, R.A., An, K.N., Chao, Y.S. (1993) Comparison
of tibiofemoral joint forces during open-kinetic-chain and
closed-kinetic-chain
exercises. J Bone and Joint Surgery 75-A(5), 732-739.
COMMENTS: Detailed kinetic model of the knee during weight-bearing and
non-weight-bearing. Only considers hamstring and quadricep muscle forces
(EMG
activity) but offers good estimations of shear as well as compressive
tiobiofemoral forces.

Although these as well as other biomechanical models offer effective
load estimations, the best method(s) that accurately measure joint
loading are those that are
not perfected yet. There are currently some research being conducted on
methods of direct load measurements that would eventually and hopefully
replace such
biomechanical models. These include robust sensors that measure
intramuscular pressure (replacing EMG) as well as articular loading.
Such sensors could have
serious clinical implications in the future.

Well, I hope this helps a little. Good luck!

Arnel

=======================================
Arnel Aguinaldo, M.A.
Biomechanical Engineer
San Diego State University
Athletic Medicine - University of California, San Diego
San Diego, CA
email: ala@znet.com
url: http://ala.someone.net/cv99.htm


__________________________________________________ ____________________

Carol,

Predicting all three (muscle, ligament and contact forces) accurately is
not
going to be easy. I am assuming you will be measuring lower limb
kinematics, EMG and ground reaction forces and using inverse dynamics to

obtain the mentioned forces. There are just too many variables to get
an
accurate prediction. Either you have to make several broad assumptions
or
use some sort of optimization to get the results. Also, are you going
to
separate the individual ligaments (e.g. anterior and posterior cruciate)
or
are you going to lump all the capsuloligamentous structures into one
passive
soft tissue structure group? The latter makes the problem more
tractable.
The next thing to consider is whether you will be creating a subject
specific model or a generic model with anthropometric scaling of certain

parameters. Finally the location of the tibiofemoral contact point is
important to accurately model knee forces. Most models either calculate

moments about an arbitratily fixed center of flexion extension, or
cruciate
crossing point, but don't take into account the back and forth sliding
of
the knee (and the contact point). This is because traditional gait
analysis
methods are not sensitive enough to pick up tibiofemoral translations.
Fluoroscopic gait analysis can measure tibiofemoral translations but it
has
its own disadvantages. We have have been working on knee kinematics in
cadaveric models. While our model uses only the quadriceps muscle for
closed chain knee extension, it serves as a good starting place to
determine
the interactions of contact point location, ligament forces, etc. More
relevant to your question, we have also developed a total knee
replacement
tibial prosthesis instrumented with force transducers and telemetry so
it is
possible to measure the actual tibiofemoral compressive force magnitude
and
location in vivo. We believe that actually measuring the contact force
and
location will reduce the unknowns to make modeling and prediction of
muscle
and ligament forces more accurate. We have already validated cadaver
models
with this prosthesis and hope to implant it in vivo in the near future.
While the prosthesis would only provide information about contact forces
in
a patient with a total knee replacement, it can be used to validate any
or
all of the models available today and thus be useful in predicting
forces in
normal gait.

Please let me know if you want any more information.

Darryl D'Lima, MD
Head, Joint Mechanics Laboratory
Scripps Clinic, CA

---------------------------------------------------------------
To unsubscribe send SIGNOFF BIOMCH-L to LISTSERV@nic.surfnet.nl
For information and archives: http://isb.ri.ccf.org/biomch-l
---------------------------------------------------------------