Test–Retest Reproducibility of Reduced-Field-of-View Density-Weighted CRT MRSI at 3T

Round 1

Reviewer 1 Report

In this paper, the Authors propose to evaluate the within- and between-subject reproducibility of a reduced-Field-of-View Density-Weighted Concentric-Ring-Trajectory MR spectroscopic imaging (rFOV-DW-CRT MRSI) protocol at 3T. This accelerated 2D Zoom approach could be of great clinical value and has already been presented in a recent MRM paper (DOI: 10.1002/mrm.28614) by the same group. A small study is conducted with only 4 healthy subjects examined twice the same day.

The manuscript is well written. However, I find the significance of this study to be quite limited. Before considering its publication, the following points should be addressed:

Q1. The number of subject is quite low. While I am not an expert in biostatistics, I would expect a larger cohort size, number of examinations, especially considering that for Subject #2, there is almost no overlap between the “test” and “retest” sessions.

Q2. There is no “gold standard” method/acquisitions to serve as a reference. It would have been enlightening to compare the CV observed for the proposed rFOV-DW-CRT MRSI protocol to a longer more conventional acquisition (for example without the acceleration provided by the non-Cartesian k-space sampling).

Q3. The aforementioned acceleration allows to achieve quite high (in-plane) nominal spatial resolution. However this choice is bound to yield spectra with reduced Signal-to-Noise Ratio (SNR). As ROI-based analysis are ultimately performed, it would have been interesting to explore alternative protocols with lower spatial resolutions but higher SNR (while still being smaller than the PCC and SCR). A discussion of this very practical aspect would be welcomed.

Q4. CV of about 10% or above for tCho, Glx and Ins are quite disappointing. Could the main causes for these variabilities be identified (B₀ shimming; low SNR; voxel positioning)?

Q5. What are the Chemical Shift Displacement Artefacts for every neurometabolite of interest? Also, one can observe undesirable signal variations at the borders of the VOI (clearly visible in Figure 1 between the left and right side of the VOI for tNAA). Were the impacted pixels excluded? If so, what were the criteria?

Q6. In addition to the metabolic maps shown in Figure 1, it would be of interest to illustrate the quality of few individual spectra and of their spectral decomposition using LCModel.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The manuscript titled “Test-Retest Reproducibility of Reduced Field of View Density-Weighted CRT MRSI at 3T” offers a thorough investigation into the reproducibility of a novel Magnetic Resonance Spectroscopic Imaging (MRSI) protocol, specifically the reduced-Field-of-View-Density-Weighted-Concentric-Ring-Trajectory (rFOV-DW-CRT) MRSI. The primary goals of this study are to evaluate the within-subject and between-subject reproducibility of the rFOV-DW-CRT MRSI protocol and to examine its effectiveness in yielding consistent results. The study reports a Coefficient of Variance for between-subject and within-subject reproducibility around or below 15% for Glx, tCho, and Myo-Ins, and below 5% for tNAA and tCr. Given the well-designed study and the organized presentation of the manuscript, I recommend accepting the manuscript for publication in the Tomography journal, contingent upon addressing the following major and minor concerns:

1. Incorporating a power analysis to determine the appropriate sample size for the study is crucial, especially given that the current sample size is only four. I suggest increasing the sample size to strengthen the conclusions.
2. Providing more detailed information regarding the selection criteria for the four volunteer subjects, including age, gender, and other pertinent demographic details and incorporating them in the statistical analysis would improve the understanding of the study's generalizability.
3. To enhance the generalizability of the findings, consider incorporating additional regions of interest (ROIs) in the study. This would allow for a broader assessment of the technique's reproducibility across different brain areas with different concentrations of the metabolites of interest.
4. The manuscript should discuss potential sources of data variability, such as subject motion or scanner instability, and elaborate in the discussion section on how these factors were mitigated during the study.
5. The manuscript would benefit from including more comprehensive details about the data processing techniques employed, specifically the parameters used for the NUFFT reconstruction and the LCModel quantification.
6. Consider adding other measures of reproducibility, such as the Repeatability coefficient and Bland-Altman analysis, to provide a more robust evaluation of the technique's reliability.
7. A more detailed discussion regarding the study's limitations and potential avenues for future research would be beneficial to understand the challenges with clinical implementation of the method.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf