See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/6534115

Development of automatic immediate fitting socket system for artificial leg

socket modeling using fuzzy control

Article  in  Conference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society.
Conference · February 2004
DOI: 10.1109/IEMBS.2004.1404369 · Source: PubMed

CITATIONS READS

3 58

3 authors, including:

Kazuhiko Sasaki
Mahidol University
23 PUBLICATIONS   30 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Smart Space Technology for Sustainable Society View project

Symbio-information Processing project View project

All content following this page was uploaded by Kazuhiko Sasaki on 22 June 2017.

The user has requested enhancement of the downloaded file.

 Development of Automatic Immediate Fitting Socket System for Artificial Leg
Socket Modeling using Fuzzy Control
Kouhei Nakamura1 , Kazuhiko Sasaki1, 2 and Kinya Fujita１
1
Tokyo University of Agriculture and Technology, JAPAN
2
National Rehabilitation Center for the Disabled, Research Institute, JAPAN
Abstract- Automatic Immediate Fitting Socket (AIFS) system II. PRINCIPLE AND CONFIGULATION OF AIFS
was developed to shorten the long term for making of an
artificial leg, by applying powder technology. AIFS is A. Principle of rigidity control
constructed of an outer shell, a particle containing bag, whose
rigidity can be controlled by the applied negative air pressure An airtight bag, which contains small particles inside, is
and three airbags between the outer shell and the particle bag stiffened by applying negative air pressure. The rigidity of
to adjust the shape by controlling the compression. A Fuzzy
controller was applied for the real-time automatic control of
the particle containing bag is affected by the void fraction,
the shape based on the subjective evaluation of the amputee. the particle diameter and the compression force as described
The surface contact pressure was evaluated during AIFS in Rumpf’s equation(3).
σ：Rigidity of Particle Bag
walking as well as the walking with the conventional socket.
The peak pressure at the Tibia end was 30 kPa in AIFS 1 -ε H ε：Void Fraction of Particles
σ  D ：Particle Diameter
walking while 90kPa with conventional socket. The pressure ε D2
distribution among Medial condyle, Fibula head and Tibia end H ：Compression Force
was more uniform with AIFS. The shape modeling The variable rigidity bag, which is controlled by changing
performance was evaluated by measuring the circumferences the negative air pressure of the airtight particle containing in
of four amputees’ stumps. The modeling error was 6%. The
shape modeling including test walking was completed in 10
this study, is the desirable property for stump shape
minutes in all four amputees and practical shapes were modeling device, because the higher rigidity allows body
obtained. weight bearing and test walking, and the lower rigidity
Keywords- Artificial leg, Shape modeling, Fuzzy control , offers the ability of shape modification. AIFS has been
Powder technology developed based on this variable rigidity.

I. INTRODUCTION B. Structure of the stump casting device

Making of an artificial leg takes long term because of
the repetition of the test walking and the socket shape
modification. One reason of the repeated shape modification
is the lack of the method for the exact stump shape modeling
during walking. The laser scanning CAD/CAM systems for
accurate stump modeling have been developed (1). However,
the accurate stump model under non-load condition does not
necessarily provide the suitable shape for walking. A tulip
limp socket system had been proposed that a flat particle
containing bag is wrapped around the stump, stiffened the
bag by applying negative air pressure, and connected to a
artificial leg parts using a flower petal shape part (2). The
tulip limb system attained early stage rehabilitation,
however, the shape of the wrapped flat bag was not
sufficient as a model of the stump. Therefore, it has been
required to develop a stump modeling system, which allows
the immediate test walking and loaded condition modeling.
In this study, we developed an Automatic Immediate Fitting
Socket (AIFS) system using Fuzzy logic control theory and
performed the clinical evaluation of AIFS.
The stump casting device consists of an elastic inner
sheath, a hard outer shell, and 0.3-0.6 mm ceramic particles
between the shell and the sheath as shown in figure1. The
inner sheaths, which contacts with the stump, is required to
be flexible for the fine shape modeling. It was made of high-
elastic rubber to allow the sheath adhere to the stump. On
⑤ ①
②
④ ⑥
⑥
③
Fig.1 Cross section of the casting device
①Inner elastic sheath ②Outer shell (acryl)
③to Vacuum pump ④EVA air tight bag
⑤Ceramic particles ⑥to Compressor
the other hand, the outer shell needs to be rigid to attain the
higher rigidity for supporting the body weight. It was made
of acrylic resin.
AIFS has three airtight compression bags made of EVA
(Ethylene Vinyl Acetate copolymer) at the inside of the
outer shell for fine shape adjustment. The upper bag is to
adjust the tightness around knee joint. The middle is to
adjust the tightness at the calf. The bottom is for the Tibia
end. The pressure of the compression bags was controlled
individually. The negative air pressure of the particle bag
was decreased to -5 kPa to reduce the rigidity for shape
modification, and increased to -80 kPa to allow test walking.
The test walking and the shape modification were repeated
until the suitable tightness was obtained.
C. System configuration
The configuration of the whole AIFS system is shown in
figure 2. The system has a compressor for the compression
bags pressurization, three electrical pressure regulators
(CKD, Type: EVS100,Scope of pressure control:0 to 98kPa)
for the control of each airbag, a vacuum pump for the
rigidity control of the particle bag, and a computer as a
controller. Moreover, there are three contact pressure
sensors (Tekscan, Type: FlexiForce, A101-1, Max load:
4.4N, Thickness: 0.1mm), which detect the contact pressure
at Medial condyle, Fibula head and Tibia end. Three air
pressure sensors (NaiS, Type:PF pressure sensor ADP1141,
Rated pressure: 98.1kPa), which detect the pressure in the
air bags were also implemented. AIFS adapts its shape to the
stump by changing the air pressure of the compression bag
by a fuzzy controller. The measured air pressure is utilized
by the fuzzy controller for the fine shape adaptation.
Fig.2 Configuration of AIFS system
D. Control algorithm of AIFS
AIFS system utilizes the subjective evaluation values
about the tightness in addition to the airbag pressure as
shown in figure 3, because the subjective evaluation of the
amputee is the most important for clinical usage. A fuzzy
controller, which is capable to handle qualitative values, was
employed in AIFS system to reflect the subjective
evaluation value of the amputee.
The input signals of the constructed fuzzy controller are
the three subjective values about the tightness at upper,
middle and bottom part of the stump and the pressure at the
three compression bags. Each input signal was fuzzificated
into three fuzzy spaces. The subjective evaluation scale is
shown in figure 4.
The fuzzy rule is a heuristic rule that was constructed
through the experience of the manual adaptation of AIFS by
an experienced Prosthetists and Orthotists who is the one of
the project members. The output of the controller was
finally defuzzied using Min-Max method and the air
pressure is controlled.
Fuzzy Controller Subjective evaluation of tightness
Loose Good Tight
Pressure of
Airbag
Fuzzy rule Defuzzy
IFS
Subject
(Airbag)
Low Middle High
Pressure of Airbag
Fig.3 Fuzzy controller of AIFS system
Loose Little loose Good Little tight Tight
0 25 50 75 100
Fig.4 Subjective evaluation scale
E. Process of stump modeling
The process of the stump shape modeling is as follows; 1)
Three contact pressure sensors are placed at Medial condyle,
Fibula head and Tibia end, where a trouble occurs
frequently, as shown in figure 5(a). 2) The inner rubber
sheath is worn on the stump as figure 5(b). 3) The stump is
inserted into the outer shell (Fig.5c). Particles are filled up
between the inner sheath and the outer shell. The shank parts,
the artificial foot and AIFS are connected and aligned at the
adequate posture. The positive air pressure of the
compression bag is adjusted to obtain the adequate contact
pressure between the stump and AIFS. 4) The maximum (-
80 kPa) negative air pressure is applied to the particle bag to
stiffen AIFS. The amputee performs test walking as figure
5(d). 5) The subjective evaluation by the amputee is scaled
and fuzzificated. The desirable compression pressure is
calculated by the fuzzy controller. The compression pressure
is automatically controlled after the negative air pressure is
decreased to –5kPa to allow shape modification. 6) The test
walking is performed again after AIFS is stiffened. The
optimal shape is obtained by repeating the process 5 and 6.
7) After the stump shape is modeled by the repetition of the
test walking and shape modification, the stump is removed
from AIFS. An impression material is poured into AIFS,
which copies the shape of the stump. Finally, a resin socket
for daily use is made from the copied model.
It turns out that improving the subjective evaluation about
tightness also improves the contact pressure disproportion.
In order to reduce the pressure at Tibia end, it is required
to attain the sufficient support at the calf before the bottom
shape adjustment, to avoid the stump drop into AIFS. In this
study, the fuzzy rules were designed to satisfy this
requirement. Figure 6(a) and (b) show that the convergence
of the pressure at upper and middle compression bags and
the tightness at knee and calf are earlier than the
convergence of the lower bag pressure and the stump end
tightness. It was confirmed that that the fuzzy controller
adjusted the air pressure and tightness correctly in the
designed manner.
Knee Joint Calf Stump End
100

Subjective Evaluation of Tightness

(Tight)

75

50 (Good)

(a) (b)
25

(Loose)
0
0 20 40 60 80 100 120 140 160 180
Time [sec]
(a)
120
Pressure of Air Bag [kPa]

(c) (d) 100

Lower Air Bag
Fig.5 Fitting procedure of AIFS system 80
(a) Attach the contact pressure sensors
(b) Wear inner elastic sheath 60

(c) Put the stump in outer shell (d) Test walk

40
Middle Air Bag
20

III. EVALUATION OF AIFS Upper Air Bag

0
0 20 40 60 80 100 120 140 160 180
A. Example of shape modeling using AIFS Time [sec]
(b)
350
Figure 6 shows an example of a shape modeling using pre Tibia End : Pre-Adjust
300
Contact Pressure [kPa]

AIFS system. The test walking was repeated nine times. The : Test Walk
shape was adjusted eight times according to the subjective 250

evaluation about tightness during test walking. Figure 6(a) 200

Medial Condyle
shows the subjective evaluation by the amputee. The 150
evaluation of each part converged to “good” (50), as Fibula Head
100
expected. Figure 6(b) shows the pressure of the compression
bags controlled by the fuzzy controller. The air pressure also 50

converged to the constant values as the subjective evaluation 0

0 20 40 60 80 100 120 140 160 180
did. It is confirmed that the fuzzy controller is capable to
Time [sec]
adjust the socket shape and the tightness automatically.
(C)
Figure 6(c) shows the contact pressure at Medial condyle,
Fibula head and Tibia end. It is observed that the pressure at Fig.6 Example of shape modeling using AIFS
Tibia end was decreased and the pressure at other points was (a) Subjective evaluation about tightness
increased. The final peak pressure at Tibia end was 40 kPa. (b) Pressure of the compression bags
(c) Contact pressure
 B. Contact pressure evaluation 16 +/- 8.7 mm smaller than the amputees’ stumps at around
the knee joint. In other words, the compression bag of AIFS
The contact pressure during walking with AIFS was reduced the circumferences 2-6%, because the soft tissue is
compared to it with a conventional resin socket as shown in easily compressed. The several percent of compression is a
figure 7 (a) and (b). The Tibia end contact pressure with common technique in the conventional casting using plaster
AIFS was 30kPa while it with the conventional socket was bandage. It appears adequate that AIFS showed the several
90 kPa. The capability of AIFS to reduce the excessive percent of compression.
contact pressure at Tibia end was demonstrated. The shape modeling including test walking with AIFS
On the other hand, the contact pressure at Medial condyle was completed in 10 minutes in all four amputees and
with AIFS was 30kPa greater than it with the conventional satisfactory shapes were obtained.
socket. The peak pressure at Tibia end was 2 times greater
than it at Medial condyle with AIFS, while the ratio with Table. 1 Subjects Profile
conventional socket is 5 times. In terms of the pressure Age Course Height (cm) Weight (kg)
distribution, it was suggested that AIFS provides more Subject. A 74 Tumor 165 80
uniform pressure distribution. Subject. B 38 Accident 176 58
Subject. C 46 Accident 175 65
100 Subject. D 61 Accident 158 50
80 400 60
Contact Pressure[kPa]

60 300 40

Circumference (mm)

Error (mm)
40 200 20

20 100 0

0 0 -20
0 2 4 6 8 10 0 60 120 180
-20 Displacement from stump end (mm)
Time[sec]
(a)
Medial Condyle Fibule Head Tibia End
(a) Fig.8 Circumference of stump and copied model
100
IV. CONCLUSION
80
We developed the Automatic Immediate Fitting Socket
Contact Pressure [kPa]

60 system, which enables immediate test walking and the

immediate socket shape modification using fuzzy logic
40 controller, which reflects the subjective evaluation about
tightness. The shape adaptation performance of AIFS was
20 evaluated in four below-knee amputees. The test walking
0
and shape modeling was successfully performed in ten
0 2 4 6 8
minutes. The more uniform contact pressure distribution
-20 was obtained. It was demonstrated that AIFS is clinically
Time [sec]
highly practical.
Medial Condyle Fibule Head Tibia End REFERENCES
(b)
Fig.7 Contact pressure during walking [1] E．D．LEMAIRE: Validation of a quantitative method
(a) with AIFS for defining CAD/CAM socket modifications. Prosthetics
(b) with conventional resin socket and Orthotics International．23 (1999) ， 30-44
[2] E．Liedberg， Tolerance of early walking with total
contact among below knee amputee． The journal of the
C. Evaluation of casting performance
international society for prosthetics and orthotics ，7，
Clinical practicality of AIFS was evaluated in four below No2，(1983) 91－95 ．
knee amputees. Table 4 shows the profiles of the subjects. [3] Edit by The Society of Powder Technology, Powder
Figure 8 shows example of the circumference of the stump technology handbook second edition. THE NIKKAN
and the cast model copied by AIFS. The copied models were KOGYO SHIMBUN, LTD. (2000), 39-46. (in Japanese)

View publication stats