Artificial Organs

35(4):404410, Wiley Periodicals, Inc.

2011, Copyright the Authors
Artificial Organs 2011, International Center for Artificial Organs and Transplantation and Wiley Periodicals, Inc.

Use of a Genetic Algorithm for Multiobjective Design

Optimization of the Femoral Stem of a Cemented Total
Hip Arthroplasty
*Toshimasa Ishida, *Ikuya Nishimura, Hiromasa Tanino, Masaru Higa, Hiroshi Ito, and
Yoshinori Mitamura
*Graduate School of Information Schience and Technology, Hokkaido University, Sapporo; Department of Orthopaedic
Surgery, Asahikawa Medical College, Asahikawa; Graduate School of Engineering, University of Hyogo, Himeji; and
School of Biological Science and Engineering, Tokai University, Sapporo, Japan

Abstract: There are many designs of the femoral stem of a

cemented total hip arthroplasty, and mechanical failure of
the stem is caused by several factors related to the cement,
such as failure of the cement. Optimization of the shape of
the stem, especially multiobjective optimization, is required
to solve these design problems because a cement fracture is
caused by multiple factors. The objective of this study was
to determine a stem geometry considering multiple factors
at the same time. A three-dimensional finite element model
of the proximal femur was developed from a composite
femur. A total of four objective functionstwo objective
functions, the largest maximum principal stress of proximal
and distal sections in the cement mantle, for each of the two
boundary conditions, walking and stair climbingwere
used. The neighborhood cultivation genetic algorithm was

introduced to minimize these objective functions. The

results showed that the geometry that leads to a decrease in
the proximal cement stress and the geometry that leads to
a decrease in the distal cement stress were not the same.
However, the results of the walking and the stair climbing
conditions matched. Five dominant stem designs were considered to be the Pareto solution, and one design was identified as the better design for all objective functions. It
was shown that multiobjective optimization using a genetic
algorithm may be used for optimizing the shape of the
femoral stem in order to avoid cement fracture. Key
Words: Computer analysisFinite element analysis
OptimizationTotal hip replacementFemoral stem
Cement failure.

Failure of the femoral prosthesis component of a

cemented total hip arthroplasty (THA) is often
attributed to failure of cement mantle associated with
many stress patterns in the cement (1).
There have been many changes in the design of the
femoral stem of THA, with many of these changes
being implicated in failure of the cement layer (2,3).
Many design and surgical factors may affect the
stress distribution related with aseptic loosening in
cemented femoral hip components (4,5). Therefore,
proper preclinical testing (6), including computer-

aided engineering, may help to reduce the frequency

of implant loosening. For example, finite element
analysis (FEA), design sensitivity analysis (7), and
shape optimization have been used to evaluate and
create stem designs (811).
Although optimization is a suitable technique for
developing femoral stems, there were many limitations in optimization procedure. Huiskes and Boeklagen (8) and Yoon et al. (9) reported one of the
earliest optimization studies. However, simple twodimensional analysis models were used in both
reports because of lack of computer resources in the
1980s. A three-dimensional analysis cannot obtain
dramatic parameter changes; however, these analyses
are important because of the asymmetric femoral
geometry. Katoozian et al. (10) reported the relationship between objective functions, boundary conditions, and optimum geometries. These authors

doi:10.1111/j.1525-1594.2010.01117.x
Received June 2009; revised July 2010.
Address correspondence and reprint requests to Dr. Ikuya Nishimura, Graduate School of Information Science and Technology,
Hokkaido University, Kita 14 Nishi 9, Kita-ku, Sapporo, Hokkaido
060-0814, Japan. E-mail: mura@bme.ist.hokudai.ac.jp

404

aor_1117

404..410

OPTIMIZATION OF A CEMENTED FEMORAL STEM

concluded that different optimization conditions led
to different results. Thus, the use of multiple conditions and objective functions, called multiobjective
optimization, is one of the best methods for solving
complex optimization problems. Fernandes et al. (11)
analyzed the multiobjective optimization of a twodimensional cementless stem with a weighting
method. This method was useful in easily obtaining
the optimization direction. However, when there is a
large number of parameters or when detailed solutions are sought, the calculation time increased
greatly because of the increase in the number of
parameter combinations.
One way of overcoming all of these limitations is to
use a genetic algorithm (GA). GA is an algorithm for
solving large-scale optimization problems. Most of
the optimization algorithms consider only one solution, whereas GA considers many solutions as one
set and renews the solution sets during the
calculation. Therefore, a sufficient number of firstsolution parameters and renewal of solution sets can
easily solve large and complex problems.
The objective of this study was to optimize the
design of a cemented stem considering the several
simultaneous causes for cement failure. For this
purpose, GA was introduced in the optimization
calculation.

405

FIG. 1. Finite element model of the proximal femur.

MATERIALS AND METHODS

Modeling and design parameters
A three-dimensional finite element model of
the proximal femur was developed (Fig. 1). The
femoral geometry was determined on the basis of
the CAD data of the composite femur. Next, the
parametric stem model was developed using a CAD
program (Pro/Engineer, PTC, Inc.). The initial stem
shape was based on the results of our previous
study (12), and the 10 design parameters considered
were defined on the basis of the stem cross-sections
(Fig. 2). Later on, this initial model was used to
define the objective function. The design parameters
were defined as medial-lateral (ML) width and
anterior-posterior (AP) width of the stem proximal,
middle, and distal cross-sections. In particular, in the
case of the proximal sections, the ML width was
divided into two parameters considering the medial
and lateral sides along the stem neck axis. The stem
length was fixed at 140 mm and the head center
offset, at 38 mm.
All design parameters were labeled L1 to L10,
and the range of parameters is listed in Table 1.
These parameters are subjected to a set of constraints

FIG. 2. Cross-section of the stem geometry.

Artif Organs, Vol. 35, No. 4, 2011

406

T. ISHIDA ET AL.
TABLE 1. Default values and the range of design
parameters used in stem CAD model

Proximal

ML
AP

Middle

ML
AP

Distal

ML
AP

L1
L2
L3
L4
L5
L6
L7
L8
L9
L10

Initial

Min

Max

8
14
6
6
12.6
6
6
8.8
3.5
3.5

6
12
4
4
10
4
4
6
2.5
2.5

21
15
6.5
6.5
13.6
6.5
6.5
9
4.5
4.5

All dimensions are in millimeters.

in order to ensure that the cross-section size

decreased from the middle cross-section to the distal
cross-section,

g 1 = L5 L8 < 0
g 2 = L6 L9 < 0
g 3 = L7 L10 < 0.
After that, a cement mantle, with a constant thickness of 2 mm was added around the stem and a
mantle, 30-mm thick (1315), was added below the tip
of the stem. This model was fed into the FEA
program (ANSYS6.0, ANSYS, Inc., Cybernet
Systems Co. Ltd., Tokyo, Japan) in the IGES format.
Static stress analysis
The values of all objective functions were calculated using a static stress analysis. All analysis models
and design parameters were automatically constructed or determined by the optimization control-

ler and the CAD and FEA programs. The FE models

comprised 11 792 8-noded brick elements and 12 667
nodes. The number of elements was the same for
all analysis models. The elastic modulii of the stem
material, bone cement (polymenthyl methacrylate
[PMMA]), cortical bone, and cancellous bone were
210, 2.2, 17, and 1 GPa, respectively (12). For each of
these materials, Poissons ratio was taken to be 0.3.
The cementbone interface was assumed to be fully
bonded. The cementstem interface was considered
to be a Coulomb frictional interface, with a coefficient of friction of 0.3 (16). We considered two
loading cases to simulate a physiological condition.
One corresponded to walking whereas the other corresponded to stair climbing. For both cases, the magnitudes and directions of the applied loads were
taken from the work of Stolk et al. (17).
Objective functions
A schematic presentation of the complete optimization procedure is shown in Fig. 3. Four objective
functions are used: the largest maximum principal
stresses of the distal (walk-distal; F1, stair-distal; F2)
and proximal (walk-proximal; F3, stair-proximal; F4)
part in the cement mantle. These objective functions
were defined on the basis of the results of the static
stress analysis of the initial model. Therefore, all optimization problems can be formulated as follows:

Minimize Fi ( x) i = 1 . . . 4
such that

(Lj )min < x j < (Lj )max g k (x) < 0 k = 1 . . . 3

FIG. 3. Procedure for design optimization.

Artif Organs, Vol. 35, No. 4, 2011

OPTIMIZATION OF A CEMENTED FEMORAL STEM

407

where Fi(x) is the objective function and (Lj)min and

(Lj)max are the upper and lower limits of the design
parameter xj.
GA parameter settings
The neighborhood cultivation genetic algorithm
(NCGA) (18) was introduced to minimize these
objective functions. NCGA is an example of a multiobjective GA and can optimize several objective
functions while maintaining the diversity of solutions.
The GA requires the encoding of design parameters
to bit strings in order to obtain the gene parameters.
In this study, it was mounted as follows:
1 Minimum value of a design parameter is defined as
000 . . . 0 on a gene parameter.
2 Maximum value is defined as 111 . . . 1.
3 A design parameter is encoded to the nearest bitstring area that is defined in (1) and (2).
A 10-bit string was used for defining the design
parameter used in this calculation. The population
size was set to 50 in a single generation, and the
number of generations was set to 20. Therefore, a
total of 1000 calculations were performed. During
each successive generation, 50% of the existing population was selected to determine the next generation
according to the fitness function. Double crossover
points on both parents organism strings were
selected. In NCGA, individuals who were close to
each other were chosen as the mated pair. Furthermore, mutation took place at a probability of 0.01 in
the case of all generations. After the analysis of each
generation, the values of objective functions were
plotted as a two-dimensional scatter chart.
RESULTS
Static stress analysis of pre-optimization
The distribution of the maximum principal stress
acting in the coronal midplane of the cement mantle
in the initial design is shown in Fig. 4. It was found
that the largest maximum principal stress in the
cement mantle was found near the tip of the stems for
both the walking (12.9 Mpa) and the stair-climbing
(15 Mpa) conditions, and another stress concentration was found in the proximal region of the cement
mantle. Note that these results were all bases of
objective functions.
Optimization and scatter diagram
Results of optimization are shown in Fig. 5. On all
charts, the direction of optimization is the lower left
one. On each diagram, each point has the following
properties:

FIG. 4. Maximum principal stress in the cement mantle, load

case of the walking (a) and stair-climbing (b) conditions.

1 solution number;
2 value of all design parameters L1 . . . L10, and
value of objective functions F1 . . . F4;
3 where the solution number contains the generation
data.
Therefore, these charts provide the relationship
between the objective functions.
Stem design for each Pareto front
Several Pareto solutions are shown in Fig. 5 and
Table 2. Figure 5a shows the resulting Pareto solutions in the two-dimensional objective function
spaces. These solutions were based on an approximate trade-off line of two objective functions (F1
and F2). In this chart, the strongest geometry for the
proximal region is (1). At the same time, (1) is a
relatively worse geometry on the Pareto front for
the distal region. From this scatter chart (Fig. 5a),
five dominant designs were selected that were on
the Pareto front ([1][5]). Figure 5b shows the linearity of objective functions of F2 and F3. Clearly,
these two objective functions are correlated and not
a trade-off. Figure 5c also shows the trade-off line of
two objective functions (F3 and F4). Each number
in the figure represents the same design as that in
another figure.

DISCUSSION
This study was an initial attempt to use a GA for the
multiobjective optimization of the femoral stem of a
cemented THA. Compared with a mathematical
optimization method such as the steepest-descent
Artif Organs, Vol. 35, No. 4, 2011

408

T. ISHIDA ET AL.

FIG. 5. Two-dimensional scatter chart of

optimization results. (a) The relationship
between walk-distal and walk-proximal, (b)
walk-distal and stair-distal, and (c) stairdistal and stair-proximal.

method, a searching method such as the GA has an

advantage for complex multiparameter analysis
because it searches the optimum values without using
a derivative function. Therefore, the GA will never
converge to a local extreme, and the solutions will
cover a wide range. Searching methods, however,
could not be used until recently, because the calculation time required to obtain good solutions was very

long. Even with the simplified analysis performed in

this study, a searching method required more than
2 months of CPU time to complete all the calculations
(Intel Xeon 2.4 GHz, Windows 2000 PC). An attraction of a searching method is that the individual calculation of each genetic parameter can easily be
carried out in parallel. Unlike in a mathematical optimization method, in a searching method, each gene

TABLE 2. Stem geometry of Pareto solution on (1) to (5).

Proximal

ML
AP

Middle

ML
AP

Distal

ML
AP

Walk-distal (Mpa)
Walk-proximal (Mpa)
Stair-distal (Mpa)
Stair-proximal (Mpa)

Artif Organs, Vol. 35, No. 4, 2011

L1
L2
L3
L4
L5
L6
L7
L8
L9
L10

(mm)
(mm)
(mm)
(mm)
(mm)
(mm)
(mm)
(mm)
(mm)
(mm)

(1)

(2)

(3)

(4)

(5)

8.47
14.7
5.28
6.33
13.1
6.18
5.95
8.77
3.24
2.54
12.4

10.93
12.43
4.81
6.2
13.46
5.15
6.27
6.6
4.07
2.78
12.2

18.61
14
5.1
5.1
13.2
6.47
4.63
6.66
3.3
2.66
9.3

19.45
14.73
5.12
6.39
13.48
6.31
4.67
6.6
3.43
2.55
8.6

19.45
13.99
5.78
6.42
10
6.32
4.66
6.1
3.55
2.51
7

3.3
12.8
4.9

3.5
13.7
5.4

3.7
10.6
7.6

3.8
9.9
5.5

4.9
8.5
7.5

OPTIMIZATION OF A CEMENTED FEMORAL STEM

parameter is independent from the others belonging
to the same generation. Thus, in this study, a 50-CPU
PC cluster was used to perform all calculations in less
than 2 days.
For general single objective optimization, the
optimal implant is only one particular case, the one
represented in the FE model. In most cases, because of
the variations in shape and properties, the optimal
implant will be suboptimal (19). Hence, a multiobjective approach was used for considering biomechanical
conditions.With respect to the Pareto solutions shown
in Fig. 5a, the geometries of (1) and (5) were the best
results for the two objective functions (Sp-walk and
Sd-walk). However, both geometries have a risk for
the other objective function. Therefore, geometries
(2), (3), and (4) were well balanced for the two objective functions. In contrast, in the case of the stairclimbing condition (Fig. 5c), (2) and (3) did not exist
on the Pareto front. Then, geometry (4) was the bestbalanced geometry for this analysis condition. Geometries (3) and (4) were almost the same except in the
case of the proximal cross-section. Therefore, the
dimensions of the stem of the proximal cross-section
should be carefully designed.
Resolution of the genetic parameter was one of the
important factors that affected the accuracy of the
solution. We used 10-bit strings as the genetic
parameters. The resolution of these strings was
approximately 1.5 10-2 mm in the case of the design
parameter L1 (15/210). These resolutions had to be
varied according to the accuracy of stem casting.
Low-resolution bit strings increase the calculation
speed but adversely affect the accuracy of the
solution. In the future work, the relationship between
the lengths of the genetic bit strings and the accuracy
of the solution has to be investigated.
From Fig. 5b, it is seen that the optimization direction of the two boundary conditions is the same. In
this study, the original values of all the loads were
obtained in an experimental study (20). In the case of
stair climbing, the load was higher than that in the
case of the walking condition for the purpose of
rotation. Therefore, the cement stress value in the
case of the stair-climbing condition was also higher
than that in the case of the walking condition, but the
tendency of results was the same in both cases. The
largest maximum principal stress value in the cement
mantle, which can lead to the fracture of the bulk of
the cement mantle, was higher in the case of the
stair-climbing condition than in the case of the
walking condition (Fig. 5). This tendency was
observed only in the proximal region. Therefore, the
stair-climbing condition was detrimental to the
cement around the proximal region. We suggest that

409

this is the reason for the synchronized optimization

direction.
The main limitation of this study was that the
failure criteria related to the boundary conditions.
We used the largest maximum principal stress as the
objective function. In some clinical reports, it has
been suggested that the failure of the femoral stem is
attributable not only to the failure of the cement
mantle but also to the failure of the cementimplant
interface (21,22). Another limitation was the accuracy of the finite element method model. Most properties and model shapes were derived from our
previous study (12) as a continued work. Therefore,
the model in this study did not include the latest
methods in modeling and material properties. Considering these limitations, we can propose a new optimization method in a future work.
CONCLUSION
This study is the first report of the use of a genetic
algorithm for optimizing the shape of the femoral
stem in a cemented total hip arthroplasty. The optimized shape was obtained considering multiobjective
functions. This design is not the best design for all
objective functions, but it is better than the existing
designs for all conditions because of the trade-off
between the objective functions. Stem designers
should find the method presented in this study useful
when modifying the shape of the stem.
Acknowledgment: This research was partially
supported by Engineous Japan, Inc., Yokohama,
Japan.
REFERENCES
1. Jasty M, Maloney WJ, Bragdon CR, OConnor DO, Haire T,
Harris WH. The initiation of failure in cemented femoral components of hip arthroplasties. J Bone Joint Surg 1991;73B:5518.
2. Dall D, Learmonth I, Solomon M, Miles A, Davenport J. Fracture and loosening of Charnley femoral stems. J Bone Joint
Surg 1993;75-B:25965.
3. Middleton R, Howie D, Costi K, Sharpe P. Effects of design
changes on cemented tapered femoral stem fixation. Clin
Orthop 1998;355:4756.
4. Mann KA, Bartel DL, Wright TM, Burstein AH. Coulomb
frictional interfaces in modeling cemented total hip
replacements. J Biomech 1995;28:126778.
5. Barrack RL. Early failure of modern cemented stems. J
Arthroplasty 2000;15:103650.
6. Stolk J, Verdonschot N, Cristofolini L, Toni A, Huiskes R.
Finite element and experimental models of cemented
hip joint reconstructions can produce similar bone and
cement strains in pre-clinical tests. J Biomech 2002;35:499
510.
7. Yang RJ, Choi KK, Crowninshfield RD, Brand RA. Design
sensitivity analysis: a new method for implant design and ad
Artif Organs, Vol. 35, No. 4, 2011

410

8.
9.

10.

11.

12.

13.
14.

T. ISHIDA ET AL.

comparison with parametric finite analysis. J Biomech 1984;17:

84954.
Huiskes R, Boeklagen R. Mathematical shape optimization of
hip prosthesis design. J Biomech 1989;22:793804.
Yoon SY, Jang GH, Kim YY. Shape optimal design of the
stem of a cemented hip prosthesis to minimize stress concentration in the cement layer. J Biomech 1989;22:1279
84.
Katoozian H, Davy DT. Effects of loading conditions and
objective function on three-dimensional shape optimization of
femoral components of hip endoprostheses. Med Eng Physics
2000;22:24351.
Fernandes PR, Folgardo J, Ruben RB. Shape optimization of a
cementless hip stem for a minimum of interface stress and
displacement. Comput Methods Biomech Biomed Engin
2004;7:5161.
Tanino H, Ito H, Higa M, et al. Three-dimensional computeraided design based design sensitivity analysis and shape optimization of the stem using adaptive p-method. J Biomech
2006;39:194853.
Estok DM, Orr TE, Harris WH. Factors affecting cement
strains near the tip of a cemented femoral component. J
Arthrop 1997;12-1:408.
Breusch SJ, Lukoschek M, Kreutzer J, Brocai D, Gruen
TA. Dependency of cement mantle thickness on fem-

Artif Organs, Vol. 35, No. 4, 2011

15.
16.
17.
18.
19.
20.
21.

22.

oral stem design and centralizer. J Arthrop 2001;16-5:648

57.
Mann KA, Gupta S, Race A, Miller MA, Cleary RJ, Ayers DC.
Cement microcracks in thin-mantle regions after in vitro
fatigue loading. J Arthrop 2004;19-5:60512.
Mann KA, Bartel DL, et al. Mechanical characteristics of the
stem-cement inteface. J Orthop Res 1991;9:798808.
Stolk J, Verdonschot N, Huiskes R. Stair climbing is more
detrimental to the cement in hip replacement than walking.
Clin Orthop 2002;405:294305.
Watanabe S, Hiroyasu T, et al. Neighborhood cultivation
genetic algorithm for multi-objective optimization problems.
IPSJ J 2002;43:No.SIG10 (TOM7);18398.
Huiskes R, Hollister SJ. From structure to process, from organ
to cell: recent developments of FE-analysis in orthopaedic
biomechanics. Trans ASME J Biomech Engin 1993;115:52027.
Bergmann G, Deuretzbacher G, Heller M, et al. Hip contact
forces and gait patterns from routine activities. J Biomech
2001;34:85972.
Mohler CG, Callaghan JJ, Collis DK, Johnston RC. Early loosening of the femoral component at the cement-prosthesis
interface after total hip replacement. J Bone Joint Surg
1995;77-A:131522.
Gardiner RC, Hozack WJ. Failure of the cement-bone
interface. J Bone Joint Surg 1994;76-B:4952.