List of Papers Browse by Subject Areas Author List
Abstract
Needle-based intervention is part of minimally invasive surgery
and has the benefit of allowing the reach of deep internal organ structures while limiting trauma. However, reaching good performance requires a skilled practitioner. This paper presents a needle-insertion training simulator for the liver based on the finite element method. One of the main challenges in developing realistic training simulators is to use fine meshes to represent organ deformations accurately while keeping a real-time constraint in the speed of computation to allow interactivity of the simulator. This is especially true for simulating accurately the region of the organs where the needle is inserted. In this paper, we propose the use of model order reduction to allow drastic gains in performance. To simulate accurately the liver which undergoes highly nonlinear local deformation along the needle-insertion path, we propose a new partition method for model order reduction: applied to the liver, we can perform FEM computations on a high-resolution mesh on the part in interaction with the needle while having model reduction elsewhere for greater computational performances. We show the combined methods with an interactive simulation of percutaneous needle-based interventions for tumor biopsy/ablation using patient-based anatomy.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2024/paper/2954_paper.pdf
SharedIt Link: pending
SpringerLink (DOI): pending
Supplementary Material: https://papers.miccai.org/miccai-2024/supp/2954_supp.zip
Link to the Code Repository
https://github.com/SofaDefrost/ModelOrderReduction
Link to the Dataset(s)
https://www.ircad.fr/research/data-sets/liver-segmentation-3d-ircadb-01/
BibTex
@InProceedings{Van_Towards_MICCAI2024,
author = { Vanneste, Félix and Martin, Claire and Goury, Olivier and Courtecuisse, Hadrien and Pernod, Erik and Cotin, Stéphane and Duriez, Christian},
title = { { Towards realistic needle insertion training simulator using partitioned model order reduction } },
booktitle = {proceedings of Medical Image Computing and Computer Assisted Intervention -- MICCAI 2024},
year = {2024},
publisher = {Springer Nature Switzerland},
volume = {LNCS 15006},
month = {October},
page = {pending}
}
Reviews
Review #1
- Please describe the contribution of the paper
The authors present a new FEM method using partitioned model order reduction for needle insertion simulator in liver. The method applies different resolutions of the mesh on different parts of the anatomy to improve the simulation speed while maintaining the fine details in the most related parts.
- Please list the main strengths of the paper; you should write about a novel formulation, an original way to use data, demonstration of clinical feasibility, a novel application, a particularly strong evaluation, or anything else that is a strong aspect of this work. Please provide details, for instance, if a method is novel, explain what aspect is novel and why this is interesting.
The paper is well written and easy to follow. The method is clearly described though some details can be more elaborated (see comments). The implementation seems to be feasible, and the video helps explain the results and shows that the simulation is realistic.
- Please list the main weaknesses of the paper. Please provide details, for instance, if you think a method is not novel, explain why and provide a reference to prior work.
The baseline is not well described (see comments below) so it is unclear about the quality of the improvement. Also, the introduction didn’t discuss other speed-up strategies, so it is unclear whether the results are significant compared with existed methods.
- Please rate the clarity and organization of this paper
Good
- Please comment on the reproducibility of the paper. Please be aware that providing code and data is a plus, but not a requirement for acceptance.
The submission does not mention open access to source code or data but provides a clear and detailed description of the algorithm to ensure reproducibility.
- Do you have any additional comments regarding the paper’s reproducibility?
N/A
- Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review. Pay specific attention to the different assessment criteria for the different paper categories (MIC, CAI, Clinical Translation of Methodology, Health Equity): https://conferences.miccai.org/2024/en/REVIEWER-GUIDELINES.html
- Section 2.2: which set of base functions alpha(t) is used in the implementation?
- Section 2.3: Though the local full fine model region is defined offline, what is the strategy to define the boundary? The authors can also discuss the balance between simulation quality and speed.
- Equation (6): does the equation requires the external force H from the needle?
- Section 3.1: The authors need to discuss how the mechanical parameters for the anatomy are obtained.
- Table 1: The results should also include standard deviation for error range estimation, and maximum error. It will also be helpful if the authors can discuss which part in the simulation has the highest error to guide further improvement.
- Page 7, precision comparisons: The authors explained the model reduction in previous section, but it is unclear what the difference is between fine reduced model and CRM (also CRM is not defined). The detailed methods of coarse model, coarse reduced model and fine reduced model are not discussed. For example, how many base functions are used and what is the truncation error for each model? What is the density of the vertexes in each model? Without these information it is hard to judge the results.
- What will be an acceptable error in the simulation? This can be an important factor related to whether the method is applicable in the real world, so I would like to see the authors having more discussion on this.
- Page 7, last paragraph, “we obtain an overall error reduction of 68% for an speed-up of 4.35”: this sentence is a bit confusing in the first read because the error reduction is comparing to the coarse model, and the speed-up is comparing to the fine model. The authors may want to report the speed of CM and CRM in Table 1 too. Minor:
- In equation (2), what are the definition of f_tn and p_t{n+1}? I think this is local values after linearization, but will be more understandable if the authors can show the linearization equations.
- Rate the paper on a scale of 1-6, 6 being the strongest (6-4: accept; 3-1: reject). Please use the entire range of the distribution. Spreading the score helps create a distribution for decision-making
Weak Reject — could be rejected, dependent on rebuttal (3)
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
My main concern is that the baseline methods are not well discussed in the experiments, so the improvement of the methods is unclear. More statistical evaluation is needed to show the improvement of the method. Beside the mean distance, the authors can also report more statistical analysis (such as maximum distance, standard deviation) to help the audience understand how stable the algorithm is. My another concern is that the authors do not discuss other state-of-the-art FEM acceleration methods in the field.
- Reviewer confidence
Somewhat confident (2)
- [Post rebuttal] After reading the author’s rebuttal, state your overall opinion of the paper if it has been changed
N/A
- [Post rebuttal] Please justify your decision
N/A
Review #2
- Please describe the contribution of the paper
This paper introduces a localized model order reduction for efficient Finite Element simulation of needle insertion in the liver. While model order reduction is not new, the main contribution here is to restrain the reduction to non-critical areas while maintaining high resolution in the region of interest. This method is implemented in a mix model combining 1) a non-reduced full Finite Element model where the needle is to be inserted with 2) a reduced model for the rest of the liver and surrounding organs. In comparison to the full non-reduced model, performance improve from 3.7 fps to 16.1 fps with an error limited to 2-3mm in mean. The mix model also outperforms models based on a coarser mesh, which is the common strategy to limit costs.
- Please list the main strengths of the paper; you should write about a novel formulation, an original way to use data, demonstration of clinical feasibility, a novel application, a particularly strong evaluation, or anything else that is a strong aspect of this work. Please provide details, for instance, if a method is novel, explain what aspect is novel and why this is interesting.
- the main strength is the idea of a localized model order reduction, to avoid losing local deformations which is inherent to global model order reduction
- the method is explained very clearly: first a summary of standard model order reduction (section 2.2) then the main contribution of local reduction (section 2.3)
- the experimental study is sound with a model integrating the liver, surrounding organs, a needle, and contacts between all these components
- results show a performance improvement with a loss of accuracy that remains acceptable in a simulation context
- Please list the main weaknesses of the paper. Please provide details, for instance, if you think a method is not novel, explain why and provide a reference to prior work.
- the main weakness is to me that the high resolution liver model is mechanically linear. I am not sure that the co-rotational formulation is enough to accurately represent the “highly nonlinear local deformation along the needle-insertion path” referred in the abstract
- while performance is improved for a linear liver model, will this improvement transfer to a non-linear model without a prohibitive loss of precision?
- Please rate the clarity and organization of this paper
Excellent
- Please comment on the reproducibility of the paper. Please be aware that providing code and data is a plus, but not a requirement for acceptance.
The submission does not mention open access to source code or data but provides a clear and detailed description of the algorithm to ensure reproducibility.
- Do you have any additional comments regarding the paper’s reproducibility?
None
- Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review. Pay specific attention to the different assessment criteria for the different paper categories (MIC, CAI, Clinical Translation of Methodology, Health Equity): https://conferences.miccai.org/2024/en/REVIEWER-GUIDELINES.html
- Fig.1: what are the two RID hyperredduction arrows? Just the areas where the stiffness and mass matrices are computed?
- p.5 why not referencing SOFA directly? The anonymization is supposedly applied to authors and groups only, not to open frameworks that could be used by anyone.
- p.6: “1000 tetrahedra, […], 9, 19 and 40 thousands”: thousand (no s), but you should use consistent notation here
- p.7: CRM not defined; Coarse Reduced Model?
- Table 1: since two coarse models are also studied in the precision study, I would add the CM and CRM performance as well in the table
- precision results reported on Fig.2 are global, not restricted to the liver only? Since you state “The error actually mostly comes from the large intestine and stomach since we still have the same kind of error reduction for the liver only: 65%.”, I understand you computed the errors for each organ. Then it would make sense to display errors by organ in the figure as well (maybe adding a detailed figure in a supplementary document)
- is the performance improvement statistically significant?
- Rate the paper on a scale of 1-6, 6 being the strongest (6-4: accept; 3-1: reject). Please use the entire range of the distribution. Spreading the score helps create a distribution for decision-making
Accept — should be accepted, independent of rebuttal (5)
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
- Interesting concept, clearly explained and tested, with a clear improvement of performance which is a crucial element for medical training simulators.
- important topic that should remain present in the MIC(CAI) conference
- no major flaw in the study
- Reviewer confidence
Confident but not absolutely certain (3)
- [Post rebuttal] After reading the author’s rebuttal, state your overall opinion of the paper if it has been changed
N/A
- [Post rebuttal] Please justify your decision
N/A
Review #3
- Please describe the contribution of the paper
The authors propose a novel local model order reduction method that, compared to previous approaches, allows for real time simulation of needle-insertion on the liver for training purpose. Real-time is achieved by employing standard model order reduction techniques on mesh models of organs surrounding the liver model. Instead of representing the liver fully as fine mesh, the novelty lies in the proposal of a local model order reduction for the liver. Regions that are further from the anticipated needle insertion area are reduced to improve computational efficiency, while the region of interest is represented with fine meshes, to accurately capture highly nonlinear deformations caused by the needle. This results in over four times faster computations.
- Please list the main strengths of the paper; you should write about a novel formulation, an original way to use data, demonstration of clinical feasibility, a novel application, a particularly strong evaluation, or anything else that is a strong aspect of this work. Please provide details, for instance, if a method is novel, explain what aspect is novel and why this is interesting.
- Usefulness: Real time simulations are crucial in the medical field (e.g. for surgical training).
- Transferability: The method can be applied to other simulations where accurate deformation computation is only relevant in a small area.
- Novelty: reformulating existing model order reduction equations from the field of computational mechanics to obtain a hybrid mathematical system capable of applying model order reduction locally resulting in more efficient medical simulations.
- Strong evaluation: capturing not only the computation acceleration achieved but also the extent to which the needle interaction behavior from the altered mesh deviates in comparison to a full fine mesh.
- Please list the main weaknesses of the paper. Please provide details, for instance, if you think a method is not novel, explain why and provide a reference to prior work.
- Potential Inflexibility: If understood correctly, the area of interest needs to be predefined, then the liver mesh will be reduced in an offline step based on the known position for needle insertion. The authors explain that this needs to be done for each insertion point and therefore for each scenario a specific reduced model needs to be created first. There is no information given on how long this takes.
- Limited comparison to state-of-the-art: Authors provide no comparison to other state of the art needle insertion simulators.
- Please rate the clarity and organization of this paper
Very Good
- Please comment on the reproducibility of the paper. Please be aware that providing code and data is a plus, but not a requirement for acceptance.
The submission has provided an anonymized link to the source code, dataset, or any other dependencies.
- Do you have any additional comments regarding the paper’s reproducibility?
No link to source code is given but to the datasets. Algorithms are explained.
- Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review. Pay specific attention to the different assessment criteria for the different paper categories (MIC, CAI, Clinical Translation of Methodology, Health Equity): https://conferences.miccai.org/2024/en/REVIEWER-GUIDELINES.html
- Further references to the mathematical formulations that already exist, and where to find them would be beneficial to a) clearly see which formulation is novel, b) facilitate a deeper understanding of the underlying mathematical formulations and where they come from.
- Source code would have been very beneficial to replicate the results. -Comparison to other, more recent medical simulations would be beneficial. All related resources seem to be over 5 years old. -The paper is well structured and well written. However, the conclusions should be seriously revised as it contains a few grammar and spelling mistakes (e.g. missing punctuation, “We” written with capital at the beginning etc.)
- On page 7 “which is 4.35 times faster that a total CUDA version”, change “that” to “than”. -the video in the supplementary materials was appreciated.
- Rate the paper on a scale of 1-6, 6 being the strongest (6-4: accept; 3-1: reject). Please use the entire range of the distribution. Spreading the score helps create a distribution for decision-making
Accept — should be accepted, independent of rebuttal (5)
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
The contributions of the paper are interesting and useful for the community. Researchers in the community could benefit from them and they seem transferable to other downstream simulations tasks. Please consider revising the conclusion for grammar and spelling.
- Reviewer confidence
Confident but not absolutely certain (3)
- [Post rebuttal] After reading the author’s rebuttal, state your overall opinion of the paper if it has been changed
N/A
- [Post rebuttal] Please justify your decision
N/A
Author Feedback
N/A
Meta-Review
Meta-review not available, early accepted paper.