Medical Engineering and Physics 38 (2016) 326–332

Contents lists available at ScienceDirect

Medical Engineering and Physics

journal homepage: www.elsevier.com/locate/medengphy

Design and evaluation of an actuated knee implant for postoperative

ligament imbalance correction✩
A. Collo a,∗, S. Almouahed b, P. Poignet a, C. Hamitouche b, E. Stindel b
a
LIRMM, CNRS UMR 5506 UM2, 161 rue Ada, Montpellier 34095, France
b
LaTIM, INSERM-UBO UMR 1101, CHRU Morvan, 2 av. Foch, Brest 29609, France

a r t i c l e i n f o a b s t r a c t

Article history: In Total Knee Arthroplasty (TKA), the collateral ligament tensioning stage cannot be standardised for all
Received 10 October 2015 patients and relies heavily on the surgeon’s experience and perception. Intraoperative inaccuracies are
Accepted 6 January 2016
practically unavoidable and may give rise to severe postoperative complications, leading to the need for
revision surgery already a few years after primary TKA. This work proposes a novel instrumented tibial
Keywords: component able to detect collateral ligament laxity conditions right after primary TKA and, if needed,
Instrumented knee prosthesis to compensate for them in the postoperative period. A miniaturised actuation system, designed to be
Knee ligament imbalance embedded in the tibial baseplate, was initially evaluated by means of 3D simulations and then fabricated
Actuation system as a full-scale prototype. Stability and force sensors tests carried out on a knee simulator allowed to
Miniaturised mechanical device
assess the effectiveness of the proposed design under normal working conditions and provided valuable
insights for future work and improvements.
© 2016 IPEM. Published by Elsevier Ltd. All rights reserved.

1. Introduction inevitably face with the intraoperative inaccuracies introduced by

A key achievement of Total Knee Arthroplasty (TKA) lies in in-
traoperatively setting up proper tension conditions of the two knee
collateral ligaments [1]. This stage, which directly determines the
postoperative Knee Joint (KJ) function and the implant lifespan,
cannot be standardised and is typically carried out by evaluating
variable and user-dependent loads [2]. Over the last decades, var-
ious assistive tools have been developed to help surgeons during
this crucial TKA ligament tensioning stage.
A few examples of instrumented knee distractors can be found
in the literature [3,4]. By controlling miniaturised mechanical
structures, surgeons are able to monitor important intraoperative
parameters such as tibiofemoral gaps and loads. Another type of
assistive tools is represented by instrumented tibial baseplates:
the use of embedded sensors, such as strain gauges [5], allows
surgeons to perform fine-tuning adjustments in ligament tension
during the component positioning process. Although, the optimal
balance conditions that surgeons can determine during surgery
✩ complications and, in the worst-case scenario, to revision surgery
The authors received no specific funding for this manuscript. No conflict of in-
terest, financial or other, exists to any authors listed in this manuscript.
∗
Corresponding author.
E-mail addresses: andrea.collo@lirmm.fr, a.collo@gmail.com (A. Collo), shaban.
almouahed@telecom-bretagne.eu (S. Almouahed), poignet@lirmm.fr (P. Poignet),
chafiaa.hamitouche@telecom-bretagne.eu (C. Hamitouche), eric.stindel@univ-brest.fr
(E. Stindel).
the replacement of the employed sensing devices with the perma-
nent prosthetic components.
In this context, over the last decade, the development of instru-
mented knee implants has been accepted as a way forward. For
instance, dedicated telemetry systems and microelectronics com-
ponents embedded in the prosthesis can provide unique intra-
and postoperative KJ kinematics data [2,6,7]. Besides improving the
quality of the surgical procedure itself, such in vivo data made
available to surgeons have turned out to be very helpful for clini-
cians and prosthesis designers as well.
However, no assistive tool has been developed to specifically
address the problem of postoperative ligament imbalance. The cru-
cial knee stabilisation process highly depends on the surgeon’s
experience and perception [8,9]. The morphofunctional evolution
that the prosthetic KJ normally undergoes during the rehabili-
tation period and in its aftermath may alter the collateral liga-
ment tension values set during surgery. Some studies reported that
some degree of postoperative collateral ligament laxity may be ac-
cepted [10] and even spontaneously corrected after TKA [11]. How-
ever, suboptimal balance conditions generally lead to postoperative
(whose number is steadily increasing [12]). This suggests there-
fore to consider the possibility to compensate for intraoperative
inaccuracies in the postoperative period; this can be achieved by
means of a miniaturised actuation system embedded in the knee
prosthesis.

http://dx.doi.org/10.1016/j.medengphy.2016.01.002
1350-4533/© 2016 IPEM. Published by Elsevier Ltd. All rights reserved.
 A. Collo et al. / Medical Engineering and Physics 38 (2016) 326–332
Fig. 1. A lax collateral ligament (left) can be properly re-tensioned by lifting the
tibial tray up on the corresponding side (right).
The aim of this work was to propose a novel instrumented knee
implant able to detect collateral ligament laxity conditions follow-
ing primary TKA and, if needed, to restore optimal balance condi-
tions before the occurrence of more severe complications. In the
long term, the proposed device is intended to strongly reduce the
need for revision surgery.
2. Materials
2.1. Implant description
The instrumented tibial component presented in this work fur-
ther develops the smart knee implant previously proposed by our
research team [13], which includes two sub-parts:
– for the evaluation of collateral ligament laxity conditions, four
PiezoCeramics (PC) elements embedded in the tibial baseplate
are employed as both force sensors and energy harvesters [14].
The collected data are meant to be processed and wirelessly
transmitted to the clinician’s computer by means of a self-
powered telemetry system [15]. This allows to define a proper
correcting action for restoring optimal balance conditions;
– for the detected imbalance correction, the miniaturised actua-
tion system proposed by Collo et al. [16] is embedded in the
tibial baseplate. In order to re-tension any detected lax collat-
eral ligament, the height and inclination of the tibial tray can
be adjusted on the corresponding side (Fig. 1). Such correct-
ing action aims to reproduce the local adjustments carried out
by means of spacer blocks during revision TKA [17] and tibial
osteotomy [18].
2.2. Clinical needs
It is very important to detect any postoperative instability as
early as possible; in the aftermath of the rehabilitation period, pe-
riodic follow-up visits allow the surgeon to monitor the prosthetic
KJ postoperative function. Any collateral ligament laxity condition
can be detected after analysing the data collected by four PC ele-
ments embedded in the tibial baseplate [13]. Accordingly, the sur-
geon defines an appropriate correcting action in terms of tibial tray
lateral uplift (4 mm maximum, in order not to impair the prosthe-
sis alignment to the lower limb mechanical axis [18]).
A progressive fine-tuning correction process is proposed by
means of successive stages: actuation is performed a first time and
the tibial tray is locked in the new raised position. Right after,
the PC elements data are processed again to check the updated
327
Fig. 2. The proposed Wedge-Leadscrew system.
KJ function. If needed, further refinements can be carried out until
optimal balance conditions are restored.
Only a continuous accurate monitoring of the evolution of the
prosthetic KJ stability combined to the use of the proposed implant
would allow to restore optimal balance conditions after primary
TKA, without the need for revision surgery. To the author’s knowl-
edge, no other ongoing study copes with the problem of postoper-
ative knee ligament imbalance using the same approach proposed
in this work.
2.3. Actuation system design
For sake of clarity, the description and the figures reported in
the following refer to the actuation of only one side of the tibial
platform (Fig. 2). The reader is invited to keep in mind that the
instrumented tibial component actuation is achieved by means of
two identical systems, one per each side of the tibial baseplate.
More details are provided elsewhere [16].
The miniaturised actuation system is based on the translation
of a wedge whose shape is custom-designed (30 mm wide, 10 mm
long, 4 mm high). The wedge is guided inside a rail grooved in the
tibial baseplate (Fig. 2). Via a threaded hole, the wedge is coupled
to a leadscrew (ISO M2x0.4 profile: 2 mm pitch diameter, 0.4 mm
pitch) which can only rotate without translating. As in a screw-nut
configuration, any leadscrew revolution produces a corresponding
wedge translation according to the thread pitch; herein, such as-
sembly will be addressed as a Wedge-Leadscrew (WL) system.
A mobile tibial tray is shaped so as to be positioned from above
onto the baseplate and fit the two WL systems. As depicted in
Fig. 2, without any actuation the tray is fully contained inside the
baseplate in a configuration that corresponds to that of conven-
tional tibial platforms. In its starting position, the wedge is aligned
to the center of the baseplate (null translation). As the leadscrew
rotates, the wedge translates laterally, towards the baseplate outer
border. By doing this, the wedge slides under the mobile tray and
lifts it upwards. The greater the wedge translation, the greater the
mobile tray lateral uplift.
328 A. Collo et al. / Medical Engineering and Physics 38 (2016) 326–332
Fig. 3. The proposed instrumented tibial component fabricated as a full-scale stainless steel prototype.
The leadscrew is held in place by specific supports and is rigidly
fixed to the shaft of a rotary stepper micromotor (entirely embed-
ded in the tibial baseplate). Once detected the lax collateral liga-
ment, only the microactuator on the corresponding side needs to
be powered in order to drive the related WL system.
2.4. System efficiency
The severe constraints imposed by a biocompatible environ-
ment were met by means of smart design choices: the combined
use of a wedge and a screw resulted in an efficient mechanism
able to develop great output forces in face of reduced input ef-
forts. Efficiency was deemed a key design aspect. The miniaturised
actuation system was needed to be sufficiently small in order to be
embedded in the tibial component and, at the same time, robust
enough to face with the strong KJ cyclic efforts [19,20]. The me-
chanical model of the proposed actuation system was accurately
studied by means of static force analysis. Different design expedi-
ents were discussed for the minimization of friction losses.
Due to the presence of the embedded actuation system, the tib-
ial baseplate is 9 mm thick. According to the wedge translation, the
mobile tibial tray lateral uplift varies in the range 0–3.6 mm and
linearly depends on the number of leadscrew revolutions (driven
by the microactuator). Two miniature thrust ball bearings prop-
erly fixed at the two leadscrew ends, together with the WL system
threaded profile, provide a robust passive locking system. Thus, the
mobile tibial tray can be rigidly blocked in the desired actuated po-
sition without the participation of the microactuator. The compo-
nent shape was defined so as to allow only the mobile tray lateral
uplift: no undesired rotation with respect to the baseplate is likely
to take place.
As reported by Lahuec et al. [15], the energy harvested by the
four PC elements can represent a power supply only for ultra-low
power consumption devices (such as the aforementioned telemetry
system). Therefore, the actuation system was assumed to be pow-
ered and wirelessly controlled respectively by means of inductive
coupling and radiofrequency techniques. All the microelectronics
components necessary for the microactuator power supply, control
and data transmission were assumed to be properly sealed inside
the hollow tibial stem (available volume: 50 cm3 [16]). The design
of an integrated low-power consumption system is currently under
development.
3. Methods
3.1. 3D simulations
The 3D model of the proposed tibial implant was developed un-
der PTC Creo design software. All the components were defined as
rigid bodies and modeled by means of three-dimensional tetrahe-
dral meshes for finite element analysis. The selected component
material was Titanium Vanadium Alloy (Ti-6Al-4V), which repre-
sents a conventional choice for knee prostheses [21].
The system was considered in a configuration of full actua-
tion on one side of the tibial baseplate (10 mm wedge transla-
tion, 3.6 mm mobile tibial tray lateral uplift). A compression load
L = 2600 N was defined perpendicularly to the mobile tibial tray
upper surface, simulating the peak tibiofemoral force acting inside
the prosthetic KJ during normal gait [20].
Under the aforementioned settings, static analysis simulations
were run on the actuated implant. The following two measures
were taken into account:
– the distribution of Von Mises Stress (VMS), a scalar value (unit:
MPa) commonly used to predict yielding of materials under any
user-defined loading condition [22];
– the deformation of the mobile tibial tray (unit: mm) under the
effect of the compression load L.
A buckling analysis was carried out to estimate the maximum
static load bearable by the mobile tibial tray before deformation. In
order to consider the worst-case postoperative scenario, the load L
was then simulated as an intermittent tibiofemoral load repeatedly
exerted onto the mobile tibial tray (1 Hz frequency). The number
of repeated loads (cycles) that the implant was able to bear before
failure was studied by means of a fatigue analysis.
3.2. Stability tests
Considering the extremely complex nature of KJ kinematics, it
was decided to carry out dynamic analyses by means of a knee
simulator. The employed knee simulator (ADL-Force 5 Simulator,
AMTI Force and Motion) features a wide range of motion and can
accurately replicate the KJ kinematics and physiological environ-
ment associated with daily life activities.
A full-scale prototype of the proposed tibial component was
fabricated using stainless steel (Fig. 3). Such choice was deemed
appropriate for the validation of the proposed actuation system
working principle. Upon satisfactory experimental outcomes, the
fabrication of a titanium alloy prototype was reckoned with as the
next necessary step (currently ongoing).
Mechanical tests were conducted to evaluate the stability of
the mobile tibial tray under different different physiological load-
ing conditions. Fifteen realistic walking scenarios were simu-
lated by progressively modifying the tibial component orienta-
tion on the knee simulator: various Internal-External Rotation (IER)
and Anterior-Posterior Displacement (APD) configurations were se-
lected as per ISO 14243 [20] in order to impair the homogeneous
contact among the prosthetic components and replicate subopti-
mal balance conditions. The stability test procedure consisted in
the following steps.
(1) Before starting any test, the knee simulator was configured
following the instructions provided by the manufacturer.
 A. Collo et al. / Medical Engineering and Physics 38 (2016) 326–332
Fig. 4. The positioning of the two ultra-thin contact piezoresistive force sensors onto the mobile tibial tray upper surface.
(2) The instrumented tibial component was manually set in its
starting position (null actuation: 0 mm mobile tibial tray lat-
eral uplift).
(3) The orientation of the implant on the knee simulator was set
in order to have full contact between the PolyEthylene (PE)
insert and the femoral component (fixed onto the simulator
condyle holder): such position was saved as Zero Position
(ZP) and considered as a reference configuration.
(4) 500 successive gait cycles were simulated (2600 N peak
tibiofemoral load, 1 s walking cycle [20]).
(5) IER was set to 15 degrees internal rotation; 500 gait cycles
were reproduced.
(6) IER was set to 15 degrees external rotation; 500 gait cycles
were replicated.
(7) The tibial component was oriented again in ZP and subjected
to 500 gait cycles.
(8) APD was set to 2 mm anterior translation; 500 gait cycles
were simulated.
(9) APD was set to 2 mm posterior translation; 500 gait cycles
were reproduced.
(10) The mobile tibial tray lateral uplift was manually set to
1.8 mm and steps (3) to (9) were repeated.
(11) The mobile tibial tray lateral uplift was manually set to
3.6 mm (full actuation condition) and steps (3) to (9) were
repeated.
3.3. Force sensors tests
The implementation of any sensor in the prototype should be
defined very accurately in order not to bias the test dynamics
replicated by the knee simulator. Such requirement was met by
positioning two Tekscan FlexiForce B201 sensing devices onto the
mobile tibial tray upper surface. Symmetrically aligned with re-
spect to the tray geometric center (Fig. 4), the two contact piezore-
sistive force sensors were individually calibrated and snug fit the
thin spacing between the tray and the PE insert.
The goal of this experimental stage was to evaluate the varia-
tions in the distribution of tibiofemoral loads on the mobile tibial
tray under different actuation system configurations. No standards
exist in terms of knee collateral ligament length and insertion sites,
which strongly vary among individuals [11,23,24]. Therefore, the
proposed actuation system is not intended to tension the collat-
eral ligaments to specific tension values (which are not accessible),
but rather to restore homogeneous load distribution conditions be-
tween the two mobile tibial tray compartments [13].
Any simulated laxity condition results in an uneven solicitation
of the two tibial tray compartments. On the prototype, this was
verified by evaluating the force sensors data. The laxity of the lat-
eral collateral ligament was reproduced on the knee simulator by
taking into account the IER and APD values already considered for
the stability tests. The tibial component was oriented at 15 degrees
329
internal rotation and 2 mm anterior translation, so as to impair
the full contact normally experienced at the PE insert-femoral
component interface. In order to allow for a more immediate in-
terpretation of the tibiofemoral load distribution, the unit chosen
for displaying the data collected by the two force sensors was the
solicitation percentage with respect to their total measurement
range (0–4448 N). Herein, such values will be addressed as M%
and L% respectively for the medial and the lateral force sensors.
To compensate for the simulated imbalance condition, the pro-
gressive correction procedure defined in Section 2.2 was followed.
The criterion for optimal restoration of balance conditions con-
sisted in visually matching the recorded graphical representations
of M% and L% . Due to the constraints imposed by the experimen-
tal settings, each successive system actuation step (manually per-
formed) was alternated to a series of 50 simulated gait cycles, nec-
essary to update the data collected by the force sensors.
4. Results
4.1. 3D simulations
As expected, the first static analysis revealed a VMS concentra-
tion in the regions where the mobile tibial tray thickness was re-
duced by the different cuts made to fit the presence of the two WL
systems. In particular, higher VMS occurred along the sharp edges
of the Medio-Lateral (ML) and Antero-Posterior (AP) cuts necessary
to host the microactuators and the wedges (red-coloured in Fig. 5).
The selected Titanium Vanadium Alloy offers a tensile yield
strength of 1100 MPa [25]. Such value, being nine times greater
than the maximum VMS computed by the static analysis (121 MPa,
Table 1), was considered to give a very reliable margin of safety.
The maximum material deformation was estimated to be on the
order of half a micrometre. Therefore, it was concluded that the
risk of failure of the mobile tibial tray was not a major concern.
However, the stress distribution was further improved by means of
two design expedients applied locally:
– the quantity of volume removed by the cuts was minimised by
locally increasing the mobile tibial tray thickness where possi-
ble. The medial and lateral edges of the tray were thickened by
1 mm (green-coloured in Fig. 5) while still fitting the presence
of the leadscrews. This allowed to strengthen the tray borders
and reduce the risk of deformation;
– the ML and AP sharp edges of the mobile tibial tray bottom sur-
face were smoothed by means of a round fillet (pink-coloured
in Fig. 5). On each side, due to the presence of the micromo-
tor and the wedge, the sharp edges were rounded off by 3 and
1 mm respectively for the ML and AP cuts. The opportunity to
round off other minor edges was also considered, but turned
out not to give any significant improvement in the distribution
of the VMS experienced by the implant.
330 A. Collo et al. / Medical Engineering and Physics 38 (2016) 326–332

Fig. 5. The mobile tibial tray bottom surface before (left) and after (right) the optimisation study (A, P, M, L respectively stand for Anterior, Posterior, Medial, Lateral). (For
interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Table 1
The main optimisation steps.

Measure 1st Static an. Thickening 1 mm fillet 2 mm fillet 3 mm fillet

Max VMS [MPa] 121 98 71 81 46

Max deformation [mm] 4.2 × 10−4 5.2 × 10−4 3.6 × 10−4 2.8 × 10−4 1.7 × 10−4

The static analysis carried out on the optimised 3D model

showed a more uniform stress distribution of the load L = 2600 N
onto the mobile tibial tray. The computed values of the maximum
VMS and deformation were both reduced by a factor of 2.5, with
the estimated maximum VMS value being almost 24 times lower
than the limit represented by the Titanium Vanadium Alloy tensile
yield strength. The static analyses results and the main optimisa-
tion steps are summarised in Table 1.
The mobile tibial tray in its fully uplifted position was esti-
mated to be capable of withstanding a maximum static load of
1.976 · 106 N (uniformly distributed on its surface) before expe-
riencing any permanent deformation.
According to the fatigue analysis results, the minimum lifespan
of the implant (in fully actuated configuration) was estimated as
Fig. 6. M% and L% data at the simulated imbalance condition, shown only for the
1.3 · 1018 cycles. The number of gait cycles performed by an active
last 10 simulated gait cycles.
patient during 15 years after primary TKA was estimated by Al-
mouahed [26] to be 4.7 · 107 ; with respect to such reference value,
thus, the instrumented knee implant proposed in this work should together with the support point on the baseplate non-actuated
offer a 2.7 · 1010 times longer lifespan. It was concluded that no side were definitively validated as sufficient conditions for the tray
mechanical failure of the mobile tibial tray was likely to occur un- stability.
der normal conditions of use.
4.3. Force sensors tests
4.2. Stability tests
With the actuation system in its starting position (null wedge
During the experimental stage the prototype underwent a total translation, null mobile tibial tray lateral uplift), the implant was
of 9 × 103 gait cycles equally distributed among three distinct subjected to 50 successive gait cycles (2600 N peak tibiofemoral
actuation cases (null, intermediate and full mobile tibial tray lat- load, 1 s walking cycle [20]). As expected, the simulated laxity of
eral uplift). In each of the fifteen studied cases, no uncontrolled the lateral collateral ligament resulted in an overload of the tibial
movement of the actuated implant took place. As anticipated in platform medial compartment. Fig. 6 reveals the non-homogeneous
Section 2.4, the corresponding profiles of the mobile tibial tray and solicitation of the two force sensors.
the baseplate were chosen so as to prevent any undesired rela- Four successive manual actuations (10, 15, 18 and 20 total com-
tive movement between the two components. This design choice plete leadscrew revolutions) were necessary to obtain matching
proved to be really effective: the tray did not experience any unde- M% and L% data. Fig. 7 depicts the evolution of the sensed forces
sired vibration or unexpected oscillation throughout multiple gait throughout the whole actuation process (for sake of clarity, only
cycles simulated in different configurations. the last 5 simulated gait cycles of each step are shown).
In addition to this, the full KJ motions experienced by the im- Table 2 summarizes the results of the force sensors tests. LS
plant did not produce any change in the wedge position. This con- stands for the number of complete leadscrew revolutions. The
firmed the soundness of the design considerations made by Collo terms M% and L% denote respectively the medial and lateral force
et al. [16] about the passive locking system. Contextually, the con- sensors data (M% and L% ) averaged over the 50 gait cycles sim-
tact line between the mobile tibial tray and the actuated wedge ulated after each actuation step. The term xCoP is a qualitative
 A. Collo et al. / Medical Engineering and Physics 38 (2016) 326–332 331

Fig. 7. Evolution of M% and L% data during the actuation process, shown only for the last 5 simulated gait cycles.

Table 2
Actuation process data.

LS M% [%] L% [%] xCoP [mm]

0 16.00 6.00 10.23

10 13.50 10.50 2.81
15 16.00 13.00 2.33
18 13.00 11.50 1.38
20 11.00 10.50 0.52

evaluation of the tibiofemoral loads Center of Pressure (CoP) po-

sition. Due to the experimental settings, the CoP lies on the line
connecting the two force sensors. With respect to the simple ref-
erence system illustrated in Fig. 4, the CoP coordinate xCoP (unit:
mm) was computed from the sensors location on the tibial tray
using Eq. (1):
M% (+22.5 ) + L% (−22.5 )
xCoP =
M% + L%
The key achievement of this experimental stage consisted in
assessing the effectiveness of the actuation procedure proposed
in Section 2.2. Initially, the simulated laxity of the lateral collat-
eral ligament resulted in an overload of the tibial platform medial
compartment (M% = 16%) with respect to the lateral one (L% =
6%). Throughout the four actuation steps, the progressive mobile
tibial tray lateral uplift effectively allowed to compensate for the
simulated ligament imbalance condition. By re-establishing proper
full contact conditions at the PE insert-femoral component inter-
face, the simulated KJ dynamics affected more homogeneously the
whole PE insert surface, without overloading only one tibial tray
compartment (M% = 11% and L% = 10.50%).
The data collected by the two force sensors further confirmed
the results of the stability tests. At each actuation step, the differ-
ence between M% and L% remained fairly constant throughout the
simulated gait cycles. This proved that the mobile tibial tray pro-
vided a stable support in all the actuated positions.
The mobile tibial tray lateral uplift varied from 0 to 2.88 mm;
consequently, the CoP gradually translated from the tibial tray me-
dial compartment (10.23 mm) towards the platform geometric cen-
ter (0.52 mm). Considering the variations in xCoP , the greatest dis-
tance was covered after the first 10 complete leadscrew revolutions
with a translation from 10.23 mm to 2.81 mm. The remaining ac-
tuation steps apparently had a smaller impact on the CoP posi-
tion. This aspect, as well as the total travelled distance (9.71 mm),
should be considered as an objective of further investigation.
The criterion of considering knee balance conditions as opti-
mally restored according to the visual matching of the two sensors
data limited any interpretation of the experimental outcomes to a
qualitative level. In these terms, the ratio between L% and M% was
Fig. 8. The evolution of IoB throughout the actuation process.
(1) defined as the Index of Balance (IoB) of the two force sensors av-
erage data. As depicted in Fig. 8, throughout the whole actuation
process IoB irregularly converged to the value 1, which represents
the ideal condition of identical solicitation of both tibial platform
compartments (L% = M% ). In more general terms, the existence of
a relationship between IoB and the mobile tibial tray lateral uplift
was deemed of great interest.
5. Discussion
5.1. Conclusion
The idea of embedding a miniaturised actuation system inside
a knee prosthesis requires a multidisciplinary research work that
must be supported by experimental data. The creation of an accu-
rate 3D model of the proposed implant allowed to carry out sim-
ulated static force analyses, to perform useful design optimisations
and to evaluate the component robustness properties. In order
to test the actuation system performances under normal working
conditions, a full-scale prototype of the instrumented tibial compo-
nent was fabricated and tested on a knee simulator. Stability and
force sensors tests were defined specifically to evaluate the actu-
ation system working principle. The effectiveness of the proposed
design choices was assessed under physiological KJ conditions. The
promising results of the experimental validation stage globally met
the expectations and provided positive indications for future stud-
ies and improvements.
5.2. Study limitations and perspectives
The validation of the actuation system working principle un-
der realistic KJ walking conditions was deemed a crucial step
332 A. Collo et al. / Medical Engineering and Physics 38 (2016) 326–332
before proceeding with further experimentation. Indeed, the actu-
ation system fatigue and wear properties, as well as those of the
PE insert, are a major concern and must be evaluated over millions
of simulated gait cycles. The ongoing fabrication of a titanium al-
loy prototype would allow to collect more reliable data about the
implant robustness properties. Contextually, it appears imperative
to define more clinically significant criteria for the assessment of
optimal balance conditions.
For biocompatibility reasons, all the mechanical and electron-
ics components must not enter in contact with the physiological
fluids that bath the knee articulation. To this purpose, the use of
an elastic micro-membrane could be envisaged. Properly fixed be-
tween the mobile tibial tray lateral edges and the baseplate, the
membrane would create a sealed volume for all the embedded
components. Biocompatible casings may also be employed as a
precautionary measure.
Upon implementation of the selected rotary stepper micromo-
tor [27] on the prototype, it would be possible to drive the actua-
tion system under more realistic conditions, as well as to investi-
gate the wireless power supply and control issues. Contextually, a
force sensor could be employed to evaluate the total force devel-
oped by the mobile tibial tray during actuation and to compare it
to the computed theoretical value.
Finally, a fully functional instrumented tibial component in-
cluding both the PC elements and the actuation system sub-parts
could be manufactured so as to evaluate the proposed implant as
a whole.
References
[1] Scuderi GR, Tria AJ. Knee arthroplasty handbook: techniques in total knee and
revision arthroplasty. New York, NY: Springer; 2006.
[2] Zimmermann F, Schwenninger C, Nolten U, Firmbach FP, Elfring R,
Radermacher K. A new approach to implant alignment and ligament balanc-
ing in total knee arthroplasty focussing on joint loads. Biomed Tech (Berl)
2012;57(4):283–91.
[3] Marmignon C, Leimnei A, Cinquin P. Robotized distraction device for knee re-
placement surgery. Int Congr 2004;1268:638–43.
[4] Marmignon C, Leimnei A, Lavallee S, Cinquin P. Automated hydraulic
tensor for total knee arthroplasty. Int J Med Robot Comput Assist Surg
2005;1(4):51–7.
[5] Crottet D, Maeder T, Fritschy D, Bleuler H, Nolte LP, Pappas IP. Development of
a force amplitude- and location-sensing device designed to improve the liga-
ment balancing procedure in TKA. IEEE Trans Biomed Eng 2005;52(9):1609–11.
[6] Hardwick ME, Pulido PA, D’Lima DD, Colwell CW Jr. e-Knee: the electronic
knee prosthesis. Orthop Nurs 2006;25(5):326–9.
[7] Arami A, Simoncini M, Atasoy O, Hasenkamp W, Bertsch A, Meurville E, et al.
Instrumented knee prosthesis for force and kinematics measurements. IEEE
Trans Autom Sci Eng 2013;10:615–24.
[8] Winemaker MJ. Perfect balance in total knee arthroplasty: the elusive compro-
mise. J Arthroplasty 2002;17(1):2–10.
[9] Schmidt F, Lampe F, Elfring R. Soft-tissue management in primary knee arthro-
plasty: common techniques, navigation and force-sensing devices. Z Orthop
Unfall 2007;145(5):599–607.
[10] Mayr HO, Reinhold M, Hube R, Von Roth P, Bernstein A, Suedkamp N, et al. Ro-
tational laxity and collateral ligament laxity following total knee arthroplasty
with rotating platform. Int Orthop 2014;38(7):1379–86.
[11] Sekiya H, Takatoku K, Takada H, Sasanuma H, Sugimoto N. Postoperative lat-
eral ligamentous laxity diminishes with time after tka in the varus knee. Clin
Orthop Relat Res 2009;467(6):1582–6.
[12] Kurtz S, Ong K, Lau E, Mowat F, Halpern M. Projections of primary and revision
hip and knee arthroplasty in the united states from 2005 to 2030. J Bone Joint
Surg 2007;89(4):780–5.
[13] Almouahed S, Gouriou M, Hamitouche C, Stindel E, Roux C. Design and
evaluation of instrumented smart knee implant. IEEE Trans Biomed Eng
2011;58(4):971–82.
[14] Almouahed S, Gouriou M, Hamitouche C, Stindel E, Roux C. The use of piezoce-
ramics as electrical energy harvesters within instrumented knee implant dur-
ing walking. IEEE ASME Trans Mechatron 2011;16(5):799–807.
[15] Lahuec C, Almouahed S, Arzel M, Gupta D, Hamitouche C, Jezequel M, et al.
A self-powered telemetry system to estimate the postoperative instability of a
knee implant. IEEE Trans Biomed Eng 2011;58(3):822–5.
[16] Collo A, Poignet P, Hamitouche C, Almouahed S, Stindel E. A miniaturised actu-
ation system embedded in an instrumented knee implant for postoperative lig-
ament imbalance correction. In: Proceedings of the 36th international confer-
ence on IEEE engineering in medicine and biology society (EMBC 2014); 2014.
p. 6211–14.Chicago, Illinois, USA
[17] Insall JN. Total knee replacement. In: Insall JN, editor. Surgery of The Knee.
Philadelphia, MS: Churchill Livingstone; 1984.
[18] Wright JM, Crockett HC, Slawski DP, Madsen MW, Windsor RE. High tibial os-
teotomy. J Am Acad Orthop Surg 2005;13:279–89.
[19] Mundermann A, Dyrby CO, D’Lima DD, Colwell CW, Andriacchi TP. In vivo knee
loading characteristics during activities of daily living as measured by an in-
strumented total knee replacement. J Orthop Res 2008;26(9):1167–72.
[20] ISO standard 14243, implants for surgery - wear of total knee-joint prosthe-
ses - part 1: Loading and displacement parameters for wear-testing machines
with load control and corresponding environmental conditions for test. Inter-
national Organisation for Standardization; 2009.
[21] Vail TP, Lang JE. Surgical techniques and instrumentation in total knee arthro-
plasty. In: Scott WN, editor. Surgery of the knee. Philadelphia, MS: Churchill
Livingstone; 2006.
[22] Ford H, Alexander JM. Advanced Mechanics of Materials. London, United King-
dom: Longmans; 1963.
[23] Bloemker KH, Guess TM, Maletsky L, Dodd K. Computational knee ligament
modeling using experimentally determined zero-load lengths. Open Biomed
Eng J 2012;6:33–41.
[24] Delport H, Labey L, DeCorte R, Innocenti B, VanderSloten J, Bellemans J. Collat-
eral ligament strains during knee joint laxity evaluation before and after TKA.
Clin Biomech 2013;28:777–82.
[25] Matweb. Titanium vanadium alloy properties. http://www.matweb.com/search/
DataSheet; 2015. [Online; accessed 13-May-2015].
[26] Almouahed S. Study, implementation and evaluation of a prototype of self-
powered diagnostic knee implant, Brest: Telecom Bretagne; 2011. Ph.D. thesis.
[27] Faulhaber. Miniature drive systems - adm0620-2r-v2-05 and precision gear-
head 06/1 k 1024:1. http://www.micromo.com/ecatalog/page245.html and
http://www.micromo.com/ecatalog/page294.html; 2015. [Online; accessed 12-
May-2015].