結合動作分析及動態X光量測技術研究活體人工膝關節之生物力學(2/2)

Total knee arthroplasty has been the
main choice of treatment for advanced
degenerative knee osteoarthritis over the
last few decades. It is essential to
provide full functional performances to
patients on total knee replacement (TKR)
design. Due to ethical considerations
and technical limitation, direct
measurement of angles during motion
and contact points of two components in
the Total knee is difficult.
The main purposes of the project
are to develop a new technique that
integrates video-fluoroscopy systems,
motion analysis systems, forceplates and
electromyography (EMG), as well as
mathematical modeling and analysis for
in vivo study of total knee biomechanics,
providing more accurate and better
descriptions of the TKR kinematics and
kinetics during isolated joint movement
and functional activities. Subjects with
PS (posterior cruciate substituting) type
TKR in one leg and CR (posterior
cruciate retaining) type in another in the
present study offer an exceptional
opportunity for study of these two
different types of knee designs without
inter-subject variations, which is helpful
for the clarification of the function of
posterior cruciate ligament (PCL) in
TKR’s. This project will be carried out
in two parts (years).
In the first part (year) of the project,
a complete kinematic and kinetic
measurement system integrating a
video-fluoroscopy system and motion
analysis equipments, namely infrared
cameras, forceplates and EMG, was
established. Patients with both PS and
CR type TKR’s performed passive and
active isolated joint movements and
functional activities including gait and
sit-to-stand while kinematic and kinetic
data were measured. Kinematic and
kinetic analysis of the lower limb with
special attention on the interaction of the
tibiofemoral was performed.
In the second part, a new
optimization method for model-based
estimation of the three-dimensional
kinematics of TKR using dynamic
fluoroscopic data was developed.
Comparisons of the accuracy and
reliability of the new method with two
existing methods was performed based
on computer simulations and
experimental data.
Present study proposed a new
technique for improving the accuracy of
3D pose estimation and accelerate
computational process without manual
operation. All 2D real data and
projection silhouettes are warped to stick
on 3D spherical surface before the
matching process. Therefore, template
matching based on shape invariant can
be applied for perspective projection
system and can divide 6 degrees of
freedom (DOF) of model to calculate
respectively.
The results from the current study
showed that the majority of patients
experienced kinematics is not similar to
those of a normal knee. However, the
extent of lateral femoral condyle
posterior rollback and the extent of axial
rotation were less.