Merge pull request #17 from Cadaei-Yuvxvs/JOSS_edit

JOSS review minor edits

Author: Cadaei-Yuvxvs

JOSS_paper/paper.bib

@@ -98,11 +98,19 @@ @article{Wang:2013
 @article{Wolff:2018,
    author = {J. Wolff and S. Schindler and C. Lucas and A. S. Binninger and L. Weinrich and J. Schreiber and U. Hegerl and H. E. Möller and M. Leitzke and S. Geyer and P. Schönknecht},
    title = {A semi-automated algorithm for hypothalamus volumetry in 3 Tesla magnetic resonance images},
-   journal = {Psychiatry research. Neuroimaging},
+   journal = {Psychiatry Research: Neuroimaging},
    volume = {277},
    pages = {45-51},
    ISSN = {0925-4927},
    DOI = {10.1016/j.pscychresns.2018.04.007},
    year = {2018}
 }
- 
+ 
+ @article{Eckstein:2022,
+   author = {K. Eckstein},
+   title = {MriResearchTools},
+   journal = {GitHub repository},
+   publisher = {GitHub},
+   year = {2018},
+   url = {https://github.com/korbinian90/MriResearchTools.jl}
+}

JOSS_paper/paper.md

@@ -85,11 +85,11 @@ Segmentation of small structures of the brain including the hypothalamus and for
 
 ## Atlas
 
-Twenty atlases were derived from manual segmentation of the hypothalamus and fornix, conducted by two tracers familiar with the hypothalamus and fornix [@Chang:OSF]. Ten non-neurodegenerative disease participants and ten patients with ALS were selected at random from within the larger datasets of the EATT4MND and 7TEA studies for the tracing. Details of the acquisition parameters have been outlined previously [@Chang:2022].
+Twenty atlases were derived from manual segmentation of the hypothalamus and fornix, conducted by two tracers familiar with the hypothalamus and fornix [@Chang:OSF]. Ten non-neurodegenerative disease participants and ten patients with ALS were selected at random from within the larger datasets of the EATT4MND and 7TEA studies for the tracing. Details of the acquisition parameters are outlined in our atlas repository [@Chang:OSF].
 
 ## Tool
 
-A summary of the pipeline is illustrated in \autoref{fig:1}. The user can specify the contrast (T1w/T2w) of the atlases used, the field strength (3T/7T) and any pre-processing steps. `OSHy-X` utilises Joint Label Fusion (`JLF`) [@Wang:2013] from Advanced Normalization Tools (`ANTs`; v2.3.1) for the registration [@Avants:2008] of atlases and segmentation of the target image. B1+ bias field inhomogeneity correction is performed using `MriResearchTools` (v0.5.2). Denoising and cropping are performed using ANTs in Python (`ANTsPy`; v0.2.0).
+A summary of the pipeline is illustrated in \autoref{fig:1}. The user can specify the contrast (T1w/T2w) of the atlases used, the field strength (3T/7T) and any pre-processing steps. `OSHy-X` utilises Joint Label Fusion (`JLF`; [@Wang:2013]) from Advanced Normalization Tools (`ANTs`; v2.3.1) for the registration [@Avants:2008] of atlases and segmentation of the target image. B1+ bias field inhomogeneity correction is performed using `MriResearchTools` (v0.5.2; [@Eckstein:2022]). Denoising and cropping are performed using ANTs in Python (`ANTsPy`; v0.2.0).
 
 ![Pipeline overview of the OSHy-X segmentation tool. Users input a target image via an one-line command, and the pipeline produces hypothalamus and fornix labels, their volumes, and a mosaic visualisation of the segmentations. The pipeline and data are encapsulated within a Docker or Apptainer container.\label{fig:1}](../Media/OSHy-X_figure_1.png)
 
@@ -105,9 +105,9 @@ Dice overlaps (\autoref{fig:3}) and ICC (Intraclass Correlation; 2-way fixed-rat
 
 ![Dice overlaps between two raters for the left and right lobes of the hypothalamus and fornix. The median Dice’s coefficient for the left and right hypothalamus is 0.94 (0.01 IQR) and 0.96 (0.03 IQR). The median Dice’s coefficient for the left and right fornix are 0.91 (0.06 IQR) and 0.91 (0.03 IQR).\label{fig:3}](../Media/OSHy-X_figure_3.png){ width=70% }
 
-`FreeSurfer` (v7.2) also offers segmentation of both the hypothalamus [@Billot:2020] and fornix [@Fischl:2002]; however the segmentation of both structures is not performed by default using the popular `recon-all` command. Overall, we found that `JLF` has higher Dice overlaps with the manual segmentations at both 3T and 7T (\autoref{fig:4}). Similarly, Dice overlaps for the fornix are significantly higher for JLF at both 3T and 7T (\autoref{fig:4}).  Additionally, we found that compared to cropped priors, whole-brain priors for `JLF` offers modest benefits to segmentation accuracy at 3T and 7T field strengths. While whole brain instead of cropped priors for `JLF` improves segmentation performance, computational time increases prohibitively.
+`FreeSurfer` (v7.2) also offers segmentation of both the hypothalamus [@Billot:2020] and fornix [@Fischl:2002]; however, the segmentation of both structures is not performed by default using the popular `recon-all` command. Overall, we found that `JLF` has higher Dice overlaps with the manual segmentations at both 3T and 7T (\autoref{fig:4}). Similarly, Dice overlaps for the fornix are significantly higher for JLF at both 3T and 7T (\autoref{fig:4}).  Additionally, we found that compared to cropped priors, whole-brain priors for `JLF` offers modest benefits to segmentation accuracy at 3T and 7T field strengths. While whole brain instead of cropped priors for `JLF` improves segmentation performance, computational time increases prohibitively.
 
-![Dice overlaps with manual segmentations for `JLF` with whole-brain priors, and FreeSurfer segmentations. For hypothalamic segmentation, median Dice's coefficients at 3T and 7T for JLF: 0.82 (0.04 IQR) and 0.83 (0.06 IQR); Freesurfer: 0.72 (0.03 IQR) and 0.72 (0.05 IQR). For fornix segmentation, median Dice's coefficients at 3T and 7T for JLF: 0.76 (0.04) and 0.77 (0.04); FreeSurfer: 0.01 (0.03) and 0.02 (0.03). For both structures at both field strengths, `JLF` outperforms the deep learning method (Wilcoxon rank sum test; \*\*\*p<0.0005 and \*\*p<0.005).\label{fig:4}](../Media/OSHy-X_figure_4.png){ width=75% }
+![Dice overlaps with manual segmentations for `JLF` with whole-brain priors, and FreeSurfer segmentations. For hypothalamic segmentation, median Dice's coefficients at 3T and 7T for JLF: 0.82 (0.04 IQR) and 0.83 (0.06 IQR); Freesurfer: 0.72 (0.03 IQR) and 0.72 (0.05 IQR). For fornix segmentation, median Dice's coefficients at 3T and 7T for JLF: 0.76 (0.04) and 0.77 (0.04); FreeSurfer: 0.01 (0.03) and 0.02 (0.03). For both structures at both field strengths, `JLF` outperforms FreeSurfer (Wilcoxon rank sum test; \*\*\*p<0.0005 and \*\*p<0.005).\label{fig:4}](../Media/OSHy-X_figure_4.png){ width=75% }
 
 # Availability