CAD-CAM Applications for Spinal
Orthotics -- Preliminary Investigation
Silvia U. Raschke, C.O.(c)
Margaret A. Bannon, B.Sc., M.Sc.
Carl G. Saunders, B.A.Sc., M. A. Sc.
William J. MeGuiness, C.P.O., (A,C), M.A.
In the summer of 1988, a joint study was
done by the Prosthetics and Orthotics Department of the British Columbia Institute of
Technology and the Medical Engineering
Resource Unit (MERU) of the University of
British Columbia. The study was undertaken
to determine the feasibility of applying existing Computer Aided Design-Computer Aided Manufacture (CAD-CAM) techniques to
the design and manufacture of spinal orthoses. The orthosis design selected was a
TLSO for the treatment of a non-structural
curve of the spine. The results of the study
were very promising. This paper describes
the study and discusses the results.
Background
The study made use of the Computer Aided Socket Design system developed at
MERU for the design and manufacture of
Trans-Tibial (TT), Below-Knee prosthetic
sockets. The system is now marketed by
Shape Technologies, Inc. under the name of
CANFIT?1,2 The objectives of the study
were:
- to determine if the modification and
milling of a torso shape was possible
using existing CAD-CAM technology,
and
- to identify those areas needing further
development to produce a system specifically for the design and manufacture
of spinal orthoses.
Methodology
In carrying out the study, we attempted to
use the CANFIT? system to modify an original torso shape to match a hand modified
torso shape. Both hand modifications and
CANFIT? modifications were done from
the same original casting of a volunteer subject.
Subject G.H. is a 10 year old male with
idiopathic scoliosis. He has a right thoracic
curve of 24° and a left lumbar curve of 23°.
He has been wearing a low profile neck ring
Cervico-Thoraco-Lumbo-Sacral orthosis
since 1986. G.H. was an ideal candidate because of his small size. As the CANFIT?
system was set up to do socket shapes, a
large torso shape would have been outside of
its range.
A conventional standing cast was taken of
the subject and a precise model of that cast
was made out of prosthetic foam. We modified the original plaster cast in the conventional manner and manufactured from it and
fitted a pelvic section to assure that the hand
modifications had been correct. We also
made a prosthetic foam model of the hand
modified cast.
Using a shape copier designed and manufactured at MERU;3 we entered the shape of
the prosthetic foam original into a data file
which could be modified in the CANFIT?
system. We also input the model of the hand
modified cast. This allowed us to superimpose the desired end result over the shape
being modified. In this way, we could check
the progress of the modification process and
direct the choice of future modifications.
The modification process consisted of
making modifications to the original shape
input and checking those modifications
against the fully modified shape which had
been input. The process was repeated until
the two matched. This method allowed close
monitoring of the ease or difficulty of matching the modifications necessary to produce a
fully modified spinal cast.
The CANFIT? system viewed the unmodified shape input as an unconventionally
shaped TT (below-knee) socket shape. The
modification functions, which allowed us to
modify the shape, were functions designed
specifically for making modifications to TT
socket shapes. Among these functions was a
"general" modification function. This function allowed us to modify less common variations on the standard socket shape (e.g.,
bone spurs, etc.).
This general modification function was the
function used to make all the modifications
on the torso shape, as the specialized TT
modifications were of no use on a torso
shape.
Once the modifications were completed,
we carved a model of the CANFIT? modified shape out of prosthetic foam with a numerically controlled milling machine. We
were interested in determining if it was possible to mill the resulting torso shape with the
manufacturing software developed for socket shapes. The results are illustrated in Figure 1
.
Discussion
The use of the general modification function to make the modifications to the torso required us to approach, in a piecemeal fashion, the modification of more complex areas.
This meant that for every modification, we
had to set the parameters of the area to be
modified, determine the location and depth
of the deepest point in the modification, and
vary the distribution of the amount of material around the deepest part.
Modifications of areas using the general modification function, therefore, consisted of a series of
"hills" (adding material) and "valleys" (removing material) layered over each other.
The time-consuming nature and the sometimes lumpy results of this method of modification were obviously not practical.
Other types of modifications, which at
present are being done both by hand and
with CANFIT? in an "add material/remove
material" manner, could be simplified and
controlled more precisely through the addition of a function which would allow blocks
of the torso shape to be rotated and shifted.
Areas in which this type of function would be
useful include:
- centering the trunk for patients with a
lateral trunk shift;
- raising or lowering a pelvic crest without disturbing its shape for a person
with an asymmetrical pelvic crest and
without disturbing its shape for a person
with a leg length discrepancy; and
- the ability to rotate the trunk over the
pelvis to derotate the thoracic spine.
Results
We concluded that modification and milling of a torso shape is possible, even though
we encountered difficulties in the process.
Modifications, such as the compression of
the abdominal area, were straightforward
and easy (Figure 2A
and Figure 2B
.) Making other
modifications, such as the buildups over the
anterior superior illiac spines (ASISs), were
possible, but more time-consuming because
positions were difficult to locate and the
modification process was not adequate.
The area which proved the most difficult
to modify was the waist crease. However, if
enough time was spent, a close approximation could be reached. Further development
in this area of milling will be required. Also,
a milling procedure which reproduces the
reverse curves over the pelvic crest and into
the waist crease is needed.
We identified one last area for development. We decided that it would be helpful to
be able to locate, on screen, the exact locations of specific reference points, such as the
ASISs, to allow for accurate placement of
modifications.
Conclusions
The areas we identified for further development will allow us to overcome the difficulties encountered in the feasibility study.
We also anticipate that this will lead to the
development of a software package specifically for the design and manufacture of spinal orthoses.
Improvements in display of landmarks,
design of reverse curves, and increased abilities to manufacture complex curves are the
components necessary to make this a clinically viable tool.
The next phase of the project will be to
develop a software and hardware package
which eliminates the problems flagged in the
feasibility study, with the end result being a
spinal CAD-CAM package similar to the
CANFIT? system.
Silvia Raschke is an Assistant Instructor and Research Orthotist at the British Columbia Institute of Technology (BCLT), 3700 Willingdon Avenue, Burnaby, British Columbia V50 3H2.
Margaret Bannon is a Research Kineseologist with the Medical Engineering Resource Unit of the University of British Columbia, Department of Orthopaedics, Vancouver, British Columbia.
Carl Saunders is Acting Director of MERU.
William McGuiness is Program Head of the BCIT School of Prosthetics and Orthotics.
References:
- Saunders, C.; J. Foort; M. Bannon; D. Dean;
and L. Panych, "Computer Aided Design of Prosthetic Sockets for Below-Knee Amputees," Prosthetics and Orthotics International, V. 9, 1985, pp.
17-22.
- Saunders, C.; M. Bannon; and J. Foort,
Computer Aided Socket Design Manual - Version 2.0. September 1985.
- Saunders, Carl, M.A. Sc., Personal Communication, April 1988. Presently Acting Director,
Medical Engineering Resource Unit, University
of British Columbia, Vancouver, British Columbia.
|
|
