Authors

Brett Levac, Sidharth Kumar, Sofia Kardonik, Jonathan I. Tamir

Abstract

Magnetic Resonance Imaging (MRI) is a widely used medical imaging modality boasting great soft tissue contrast without ionizing radiation, but unfortunately suffers from long acquisition times. Long scan times can lead to motion artifacts, for example due to bulk patient motion such as head movement and periodic motion produced by the heart or lungs. Motion artifacts can degrade image quality and in some cases render the scans nondiagnostic. To combat this problem, prospective and retrospective motion correction techniques have been introduced. More recently, data driven methods using deep neural networks have been proposed. As a large number of publicly available MRI datasets are based on Fast Spin Echo (FSE) sequences, methods that use them for training should incorporate the correct FSE acquisition dynamics. Unfortunately, when simulating training data, many approaches fail to generate accurate motion-corrupt images by neglecting the effects of the temporal ordering of the k-space lines as well as neglecting the signal decay throughout the FSE echo train. In this work, we highlight this consequence and demonstrate a training method which correctly simulates the data acquisition process of FSE sequences with higher fidelity by including sample ordering and signal decay dynamics. Through numerical experiments, we show that accounting for the FSE acquisition leads to better motion correction performance during inference.

Link to paper

DOI: https://link.springer.com/chapter/10.1007/978-3-031-16446-0_67

SharedIt: https://rdcu.be/cVRUb

Link to the code repository

https://github.com/utcsilab/GAN_motion_correction

Link to the dataset(s)

N/A

Reviews

Review #1

Please describe the contribution of the paper

The paper presented a method for simulating synthetic motion artifacts accounting for k-space acquisition ordering and T2-signal decay in FSE acquisition and showed improved motion correction
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.
1. Motion corrupted FSE images were simulated by accounting for MRI acquisition, which improves the realistic representation of training data.
2. The proposed approach was compared with the models trained by naïve motion corrupted FSE image, showing its superiorities.
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.
1. It wasn’t clear why FSE images were generated by multi-contrast brain MRI data. My biggest concern is both training and testing data were only based on simulated motion/FSE images. Testing the model onto true FSE images with actual motion corruption will provide a better understanding of how the data-driven approach would improve the motion compensation.
2. Motion artifacts were simulated by k-space ordering and signal decay, which will be highly dependent on echo train length and TR, but it seemed like one set of ETL and TR was considered.
3. How many testing datasets were used?
Please rate the clarity and organization of this paper

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

No issues.
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

FSE is often used as a multi-slice 2D FSE, and motion corruption can happen in both in-plane and through-plane directions. Any methods considered in both directions would be more practically useful.
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

5
Please justify your recommendation. What were the major factors that led you to your overall score for this paper?

The study showed importance of using realistic motion artifacts for training for motion correction approaches.
Number of papers in your stack

5
What is the ranking of this paper in your review stack?

2
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

In this work, the authors propose a FSE motion-corrupted data generation method that takes into account the intra-echo signal decay.
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.
- Well-structured
- Clearly motivated
- Neat figures
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.
- Additional training details are not listed in the supplementary material
- Reason for training a cGAN as opposed to other available models/architectures are not stated
Please rate the clarity and organization of this paper

Excellent
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

Seems reproducible since the authors have committed to sharing the code upon publication.
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

Strongly recommend the authors to bolster their work by performing experiments on prospectively motion-corrupted FSE data.
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?

Clearly motivated and well implemented; well written with neat figures; commitment to sharing code
Number of papers in your stack

5
What is the ranking of this paper in your review stack?

3
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 #3

Please describe the contribution of the paper

This paper proposes a new method for simulating motion corrupt data in FSE sequence, which can be used for training deep learning model for retrospective motion correction. The method considers the key effects of FSE sequence such as signal decay and sample ordering and can simulate more realistic motion corrupted data. Experiment results showed that by training with data simulated by the proposed method, a significant improvement can be achieved in motion correction results.
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.
1. The paper pointed out a critical problem in existing motion simulation methods in MRI, and proposed a reasonable way to solve it. The motivation is clear and sound.
2. The idea of taking into account the effects of FSE when simulating motion is insightful and interesting. The proposed method is simple yet effective, achieving superior performance for motion correction compared to conventional method.
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.
1. The experiments are lacking. Experiments were only carried out on simulated dataset. The method was not evaluated on actually acquired MRI data with real motion.
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

Good.
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
1. Evaluate the method on real MRI data with real motion.
2. Add PSNR in Table 1.
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?

Novel, interesting and insightful idea. Simple yet elegant method design. Superior performance in experimental results. However, lacks results on real data.
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

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.
In this work, the authors propose an FSE motion-corrupted data generation method that takes into account the intra-echo signal decay. While reviewers were overly positive they also indicated some important aspects that should be considered by the authors as follows:
1. Demonstration of the impact of the method on real data rather than on simulated data from the same source as the training data.
2. Quality of the simulation and its representation of real motion (i.e. inter-slice motion rather than just intra-slice motion, etc). Based on the positive reviews I recommend to accept this 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).

1

Author Feedback

We thank all the reviewers for their valuable comments regarding our work. In response to all reviewers, we aim to verify the results on prospectively acquired data by conducting scans at our institution. While for the response deadline, we do not have sufficient time to procure a significant number of subjects for in-vivo results, we plan to scan and verify with motion corrupt in-vivo data (not simulated) for the camera-ready deadline.

Reviewer 1 expressed concerns with the limited number of ETL and TR combinations we included in our result section. Using our framework for simulation it is straightforward to change simulated scan parameters and thus for the camera ready deadline we will train additional models and include results for a larger variety of FSE acquisition parameters. As for the test dataset, we used 3 subjects each with 15 2D axial slices. In regards to the concern over intra-slice motion, the current work was a proof of concept to show that there is a requirement of correct motion modeling for FSE data. Unfortunately the multi-contrast sequence that we had for estimating the parameter maps (T1, T2 & PD maps) is based on 2D multi-slice imaging which kept us from creating volumetric parameter maps. That’s why we only considered inter-slice motion simulations. To accurately simulate intra-slice motion (out-of-plane motion) we need 3D volumetric data to simulate the effects of slice selection which influences the effects of intra-slice motion. In the future, this work can be extended to intra-slice motion simulation by collecting 3D k-space to estimate tissue parameters for a volume. Thus we make no claims on intra-slice motion artifact performance.

We note reviewers 2 and 3 commenting on the lack of training and performance specifics (training epochs, learning rate, PSNR) in our manuscript and we will add this to our final paper.

back to top

FSE Compensated Motion Correction for MRI Using Data Driven Methods