List of Papers By topics Author List
| Paper Info | Reviews | Meta-review | Author Feedback | Post-Rebuttal Meta-reviews |
Authors
Shuai Zhang, Liang Zhao, Shoudong Huang, Hua Wang, Qi Luo, Qi Hao
Abstract
Total knee arthroplasty (TKA) is a common orthopaedic surgery to replace a damaged knee joint with artificial implants. The inaccuracy of achieving the planned implant position can result in the risk of implant component aseptic loosening, wear out, and even a joint revision, and those failures most of the time occur on the tibial side in the conventional jig-based TKA (CON-TKA). This study aims to precisely evaluate the accuracy of the proximal tibial resection plane intra-operatively in real-time such that the evaluation processing changes very little on the CON-TKA operative procedure. Two X-ray radiographs captured during the proximal tibial resection phase together with a preoperative patient-specific tibia 3D mesh model segmented from computed tomography (CT) scans and a trocar pin 3D mesh model are used in the proposed simultaneous localisation and mapping (SLAM) system to estimate the proximal tibial resection plane. Validations using both simulation and in-vivo datasets are performed to demonstrate the robustness and the potential clinical value of the proposed algorithm.
Link to paper
DOI: https://link.springer.com/chapter/10.1007/978-3-031-16449-1_13
SharedIt: https://rdcu.be/cVRUT
Link to the code repository
https://github.com/zsustc/Calibration
Link to the dataset(s)
https://github.com/zsustc/Calibration
Reviews
Review #1
- Please describe the contribution of the paper
The authors propose a method to evaluate the 3D alignment of the tibia resection block during a total knee arthroplasty procedure based on a series of intraoperative fluoroscopy images and a preoperative CT scan of the tibia. The author(s) tested the proposed method in a simulation and clinical pilot study.
- 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.
According to the reported results, the proposed method outperforms previous methods with regards to robustness and accuracy, while ensuring a short runtime.
- 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.
I’m doubtful about the clinical applicability of the proposed method. Although the authors tested the clinical feasibility of the method in a pilot clinical study, in my opinion, the advantage of using such system in clinical practice was not sufficiently discussed.
- Please rate the clarity and organization of this paper
Very Good
- Please comment on the reproducibility of the paper. Note, that authors have filled out a reproducibility checklist upon submission. Please be aware that authors are not required to meet all criteria on the checklist - for instance, providing code and data is a plus, but not a requirement for acceptance
Author(s) response to reproducibility indicate that data and code will be shared, which makes this paper full reproducible.
- Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review: https://conferences.miccai.org/2022/en/REVIEWER-GUIDELINES.html
Firstly, I would like to comment that the accuracy of the results you have presented in the manuscript is impressive. However, I have a few questions regarding the method used to create these results:
- From my understanding, the algorithm requires as input an initial state, comprising of an estimated C-Arm pose for each fluoroscopy image, and initial 3D pin poses. According to equation (3) the state also contains a set of 3D back-projection points for each contour edge point and each fluoroscopy image. I’m unsure how these 3D points are initialized? As mentioned in equation (1) and (2) the relationship between 2d contour points and 3D back-projected point is only proportional.
- Would the algorithm not also require the tibia and pin contour points from the fluoroscopy images? I believe in figure 3, these are named contour observations? If so, how are these contour points determined? How would the accuracy of the contour detection influence the overall results?
- In your simulation experiment (section 4.1) what was the resolution of the fluoroscopy images? The differences in your results compared to the other methods are noteworthy. Could you discuss your results and why you think that your method outperforms the comparative methods?
- The colors-dots in the legends of the images in Fig 3 are too small to identify
- According to the supplemental material the tests seem to be performed all with N=2 (number of fluoroscopy images). This information should be added to the manuscript. Also, how would change in N effect the results, i.e. N=1, N=10? Despite the good results, I have some hesitation about the clinical feasibility and applicability of the proposed method. For example, the following sentence in the manuscript “… the tibial resection plane estimation can be completed without much influence on the CON-TKA procedure.” gives the impression that the method can be applied at the cost of the runtime of the optimization. To the best of my knowledge, fluoroscopy imaging is not standard of care during a conventional TKA procedure in many clinics. Therefore, setup and obtaining the images should be considered part of the proposed method and clinical feasibility and applicability should be considered including these extra steps and additional radiation exposure. In the introduction you mentioned that your proposed method can be used as replacement of systems which are actively navigating the tibia resection block and/or tibia resection. Arguments against the computer-assisted navigation system includes: extra procedure steps, time, complexity and instruments. Although I agree that these are well established and published disadvantages of many computer-assisted TKA systems, I believe it would need to be evaluated if your proposed method can in fact eliminate these concerns. As mentioned above, procedure time and complexity might be influences by the use of fluoroscopy. Furthermore, while computer-assisted system are designed as a “one-step” navigation, your proposed method would need a repeat alignment of the tibial block when an error is detected. Lastly, you mentioned also that no significant improvement of mid-to-long term functional outcome was detected using navigation. However, in the publication you refer to, the authors did measure a significant improvement of tibial component alignment compared to conventional method. Your method also aims to improve the tibial component alignment. I would therefore be interested to get your thoughts on if and how your system might have the potential to overcome this mid-to-long term problem.
- Rate the paper on a scale of 1-8, 8 being the strongest (8-5: accept; 4-1: reject). Spreading the score helps create a distribution for decision-making
4
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
- the remaining open questions to the methods make it difficult for me to fully assess the merit of the proposed method
- Number of papers in your stack
4
- What is the ranking of this paper in your review stack?
3
- Reviewer confidence
Confident but not absolutely certain
- [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
This paper presents a real-time algorithm for reliably and intraoperatively estimating the tibial resection plane for CON-TKAs. It solves a SLAM problem using a patient-specific pre-operative tibia CT scans, a trocar pin mesh model and two intra-operative X-ray images. Simulation experiments demonstrate the robustness and accuracy of their proposed algorithm.
- 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 authors novelly formulate intraoperative tibial resection plane estimation as a SLAM problem, and their simulation results show the possibility to apply their method in the clinical settings.
- 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.
There is no clinical evaluation in the current study.
- Please rate the clarity and organization of this paper
Very Good
- Please comment on the reproducibility of the paper. Note, that authors have filled out a reproducibility checklist upon submission. Please be aware that authors are not required to meet all criteria on the checklist - for instance, providing code and data is a plus, but not a requirement for acceptance
The implementation details have been well addressed.
- Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review: https://conferences.miccai.org/2022/en/REVIEWER-GUIDELINES.html
The authors present an important method for improving the CON-TKAs by formulating the cutting plane estimation method as a SLAM method, and demonstrate the effectiveness of their method by the in-vivo experiments. Some suggestions are give below. The authors should add the clinical evaluation or add the corresponding discussion in the manuscript, and should carefully proofread their manuscript to correct the existing typos, e.g., “gold standar”.
- Rate the paper on a scale of 1-8, 8 being the strongest (8-5: accept; 4-1: reject). Spreading the score helps create a distribution for decision-making
6
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
The formulation method is interesting and critical for improving the intra-operative TKA surgeries.
- Number of papers in your stack
4
- What is the ranking of this paper in your review stack?
1
- Reviewer confidence
Very confident
- [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 proposed a SLAM based approach to accurately estimate the proximal tibial resection plane intra-operatively, using 2 X-ray radiographs, 3D tibia mesh model, and trocar pin 3D mesh model. Simulation and in-vivo experiments demonstrated its good performance.
- 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.
- Clinical significance was well illustrated. The authors clearly showed the impact of using this technique on TKA.
- The problem has been re-modeled mathematically to be solved by SLAM. The definition of different data term were well formulated.
- Validation on simulation and in-vivo experiments showed promising results.
- 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.
Proof-reading is required.
- Please rate the clarity and organization of this paper
Very Good
- Please comment on the reproducibility of the paper. Note, that authors have filled out a reproducibility checklist upon submission. Please be aware that authors are not required to meet all criteria on the checklist - for instance, providing code and data is a plus, but not a requirement for acceptance
The reproducibility depends on whether the code is to be open-sourced. Otherwise, making a SLAM pipeline working takes a lot of efforts in fine-tuning. It is good that the authors promise to share the code after acceptance.
- Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review: https://conferences.miccai.org/2022/en/REVIEWER-GUIDELINES.html
More proofreading is needed to correct the typos, e.g. “gold standar(d)”. Enlarge the legend font size in Fig.3
- Rate the paper on a scale of 1-8, 8 being the strongest (8-5: accept; 4-1: reject). Spreading the score helps create a distribution for decision-making
7
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
The paper is very well written. I also like the elegant solution: SLAM is not novel but the application of SLAM on TKA seems novel and the paper demonstrated promising accuracy and robustness in terms of resection plane estimation (<3 degree).
- Number of papers in your stack
4
- What is the ranking of this paper in your review stack?
1
- Reviewer confidence
Confident but not absolutely certain
- [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 #4
- Please describe the contribution of the paper
This manuscript proposes a method to align the 2D points in the X-ray images and 3D points in the reconstructed mesh by leveraging the idea of SLAM. The alignment (pose estimation) is accurate enough such that the obtained resection planes have smaller angle errors.
- 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 problem is important and this paper proposes a sound algorithm to achieve considerable performance. Compared with previous works, the proposed framework is more robust.
-
The manuscript is well written and easy to follow.
-
- 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.
- Some implementation details should be included:
- Page 3, the second paragraph, how to obtain those edge pixels?
- How to obtain the GT plane?
- Page 7, the last paragraph, the authors mentions that the proposed algorithm requires 0.5-1 minimutes to converge. Compared with other methods, the processing time is short, but be careful using ‘real-time’.
- Some implementation details should be included:
- Please rate the clarity and organization of this paper
Very Good
- Please comment on the reproducibility of the paper. Note, that authors have filled out a reproducibility checklist upon submission. Please be aware that authors are not required to meet all criteria on the checklist - for instance, providing code and data is a plus, but not a requirement for acceptance
The authors should provide more implementation details or release the code.
- Please provide detailed and constructive comments for the authors. Please also refer to our Reviewer’s guide on what makes a good review: https://conferences.miccai.org/2022/en/REVIEWER-GUIDELINES.html
This manuscript is well written, the ideas and methodology are well structured and presented. The proposed algorithm is robust and has good performance. Some subscripts in Section 2 and 3.1 can be removed for better understanding. Be careful using ‘real-time’.
- Rate the paper on a scale of 1-8, 8 being the strongest (8-5: accept; 4-1: reject). Spreading the score helps create a distribution for decision-making
6
- Please justify your recommendation. What were the major factors that led you to your overall score for this paper?
The contribution to the community is considerable. The algorithm is straightforward but achieves good performance.
- Number of papers in your stack
1
- What is the ranking of this paper in your review stack?
1
- Reviewer confidence
Confident but not absolutely certain
- [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
Primary Meta-Review
- Please provide your assessment of this work, taking into account all reviews. Summarize the key strengths and weaknesses of the paper and justify your recommendation. In case you deviate from the reviewers’ recommendations, explain in detail the reasons why. In case of an invitation for rebuttal, clarify which points are important to address in the rebuttal.
Strength
- SLAM-based solution for estimating the tibia resection block during a total knee arthroplasty procedure based on intra-operatively acquired C-arm images, a CT scan of the tibia and pin models.
- Conducting simulation and in-vivo experiments to demonstrate the feasibility
Weakness
- Overstated claims as pointed out by reviewer #1 and #4
- As pointed out by Reviewer #1, how to initialize the algorithm, including pose initialization for tibia model as well as pin models? And how to extract edge points?
- The influence of C-arm image segmentation errors on the overall performance
- Proof-reading the paper
- What is the ranking of this paper in your stack? Use a number between 1 (best paper in your stack) and n (worst paper in your stack of n papers). If this paper is among the bottom 30% of your stack, feel free to use NR (not ranked).
2
Author Feedback
We thank Reviewer #1 (R1), R2, R3, R4 and Area Chair (AC) for reviewing our paper. They appreciate that we are addressing “an important problem” (R3, R4), “novelly formulate intraoperative tibial resection plane estimation as a SLAM problem” (R2, R3), “the paper demonstrated promising accuracy, robustness and effectiveness in terms of resection plane estimation” (R2), “demonstrate the effectiveness of their method by the in-vivo experiments.” (R3), “the contribution to the community is considerable.”(R4).
We now address the comments by AC and the “major weaknesses” raised by R1:
AC and R1 would like to know how to initialize the proposed algorithm. The X-ray frame poses were initialised by the X-ray C-arm rotation angles and translation measured using the device joint encoders, and the pin poses w.r.t the pre-operative tibia model were initialised by the pose of the ideal resection plane in the pre-operative tibia model. The state variables 3D back-projection tibia contour points were initialised by back-projecting their corresponding 2D tibia contour points into the pre-operative tibia model frame using the initial X-ray frame pose and the initial guesses for the point depths. And the points depth were initialised by using the distance between the patella of the patient and the X-ray tube from the C-arm X-ray device. Similar method is used to initialise the 3D back-projection pin contour points. We will state this more clearly in the final version of the paper.
R1 asks why the proposed method outperforms the comparative methods. As stated in Section 3.1, “Overall, the three energy terms are used together to ensure the accuracy and robustness of the estimation algorithm.” In detail, the edge observations of the tibia, fibula and pins on the intraoperative X-ray images, and their corresponding pre-operative 3D mesh models are aligned in both 2D (contour re-projection term and model projection term) and 3D (contour back-projection term). The model projection term is critical for dealing with the edge feature scaling problem in which the size of the contours extracted from model projection will be smaller or larger than the observed contours in the intra-operative X-ray frames. In the contour back-projection term, we optimise the 3D x, y and z coordinates instead of directly optimising the depths. We also found that letting the 3D point parameters freely change in 3D space during each iteration has much better convergence property than restricting the changes only on the direction of the observation ray from the optical centre. We will state this more clearly in the final version of the paper.
R1 mentions that fluoroscopy imaging is not standard of care during a CON-TKA procedure. Pre-operative and post-operative X-ray imaging are essential and the most frequently used in planning and validating the TKA surgical procedure [1]. The C-arm X-ray device is always available in operation rooms and easily accessible to surgeons. In the proposed framework, two views of intra-operative X-ray frames are captured during the procedure before the surgeon cuts the proximal end of the tibia and this step normally takes less than one minute, which does not interrupt too much on the workflow of CON-TKA. This is what the surgeon did to collect in-vivo datasets for our experiments in the paper. We will state this more clearly in the final version of the paper.
Regarding the mid-to-long term evaluation (R1, R3). Currently, simulations are used to test the accuracy and robustness of the proposed algorithm, and the experiments using in-vivo datasets from five patients demonstrate the effectiveness of the proposed method. The intra-operative experiments and mid-to-long term evaluations will be our future work. We will state this more clearly in the final version of the paper.
[1] Cilengir, Atilla Hikmet, et al. “Preoperative Radiological Assessment of The Total Knee Arthroplasty.” Forbes Journal of Medicine 2.2 (2021).