|Please use this identifier to cite or link to this publication: http://hdl.handle.net/10380/1367|
(This review can be also viewed by opening the attached PDF file)
The idea of pre-operative prediction of mechanical properties of dental implant using
finite element method presented in the paper is very exciting. However, the
implementation of this idea appears to be flawed, which includes the following:
1) Justification for building “a patient-specific” model using a generic mesh and
patient-specific bone material properties is not given.
2) Selection of boundary conditions for the implant (the implant is rigidly fixed
to the bone) lacks justification.
3) Verification of in-house finite element code developed in the study is
4) The finite element modelling results grossly differ from the experimental data
obtained using artificial bone.
Therefore, I cannot recommend this manuscript for publication in its current form. A
major revision of the entire study should be undertaken.
1) It is commonly accepted practice that newly developed numerical algorithms
(such as e.g. in-house finite element codes) are at first verified against the
existing well-established numerical procedures (such as e.g. those
implemented in commercial finite element codes) or against simple examples
for which solutions are known (e.g. Viceconti et al. 2005, Zienkiewicz 2000).
The Authors should include such verification of their code in the paper.
2) Finite element method is a numerical method for solving partial differential
equations. Therefore, determining correlation coefficients between the
experimental and finite element modelling results when validating finite
element models as done in Results section is misleading. Close to linear
relationship between the modelling and experimental results is not sufficient
to claim that the modelling results are “promising … in terms of accuracy …”
as stated by the Authors in Discussion. The results shown in Figure 5 indicate
that the experimentally determined removal torque is around five times greater
than the one predicted using the finite element model. This is a clear
indication of gross disagreement between the modelling and experimental
results and evidence of serious flaw in either modelling or experimental (or
both) part of the study. Differences of 500% between the modelling and
experimental are unacceptable by any engineering standard given the fact the
constitutive properties of the modelled continuum are well-known. These
differences call for major revision of the entire study.
3) Figure 6 clearly indicates that the relationship between the modelling and
experimental results is non-linear. Justification for drawing a straight line
trough the points clearly following non-linear pattern should be given. As was
with Figure 5, 200% difference between the experimental and modelling
results should be explained, and relevant revision of the study should be
undertaken to eliminate any difference of such large magnitude.
4) P. 3: “The removal torque is approximated by multiplying the friction
coefficient by the total of the radial reaction forces due to pressfit at the
bone/implant interface”. Given the fact that no contact between the bone and
implant was modelled, how was the radial reaction force calculated?
5) Quality of the results obtained using a finite element model is determined not
only by accuracy of the constitutive parameters for the model but also by how
closely the model represents the geometry and boundary conditions of the
analysed structure (see e.g. Bathe 1996). What was the rationale for the
Authors’ assumption that patient-specific constitutive properties and generic
implant and bone geometry are sufficient to build a patient-specific boneimplant
6) How were the dimensions of the bone part of the model determined?
7) In the first paragraph of section 2.1.2 Construction of patient-specific finite
element analysis the Authors state that “The mesh used for the simulation is a
generic mesh …”. However, in the second paragraph in the same section they
write that “ Elements outside of the bone are assigned near zero …stiffness,
making the patient specific mesh construction entirely automatic”. This seems
a contradiction. Was the patient-specific or generic mesh used?
8) What do the Authors mean by the “implant stability” given the fact that they
assume that there is no movement between the bone and implant?
The paper should be re-written using the single-column format (see the Insight
Journal submission template at
Bathe, K.-J., 1996. Finite Element Procedures. Prentice-Hall, Upper Saddle River.
Viceconti, M., Olsen, S., Nolte, L.-P., and Burton, K., 2005. "Extracting clinically
relevant data from finite element simulations," Clinical Biomechanics, Vol. 20,
Zienkiewicz, O. C. and Taylor, R. L., 2000. The Finite Element Method. Vol. 2: Solid
mechanics London: McGraw-Hill.
|Keywords:||osseointegration, planning, finite element, implant|
Linked Publications more
A File Reader for the VisualSonics Vevo 770 Digital RF Data
by McCormick M.
Voxelizer plug-in for Blender
by Grothausmann R., Gout J., Kühnel M.
Send a message to the author