Micro-Computed Tomography Soft Tissue Biological Specimens Image Data Visualization

Round 1

Reviewer 1 Report

## General remarks
The manuscript describes *one* way of visualizing tomographic data by modifying/editing the look-up tables (LUTs) mapping gray values in the data to (color) representations on screen.
In the current state it fails to highlight any new scientific results and is in my opinion only a simple statement about having extensively played with the LUTs for the visualization of tomographic data.
The abstract mentions an automatic optimization of the LUTs which is *not* done in the manuscript.
The 53 mentioned transfer functions are in *no* way specified in more detail, apart from 'LUT x' was used here and 'LUT y' was used here, which fails to contribute to any knowledge gain for the authors.
The details about the LUTs specified in the 'materials and methods' section (namely Fig. 3 and Fig. 4) are in themselves quite interesting, but would - in my opinion - better be suited in a learning resource for undergraduate students starting with the process of visualizing tomographic data.

## General issues
Even thought 8 (!) authors claim to 'have read and agreed to the published version of the manuscript' it would profit from some *very* thorough proof-reading and rework.
In at least one instance, there's a repeated sentence in one paragraph, crucial preparation steps for the biological specimens are omitted, details about the tomographic data acquisition are repeated in needless detail and a large, overarching story is missing from the manuscript.

I fail to see any scientific contribution in this manuscript and would only want to read it again after the authors have fixes *several major* concerns of mine.

## Detailed comments
### Abstract
- The *first* sentence contains either a typo or is in itself not a valid sentence: either research was done to verif*y* an optimal transfer function or 'verified' does fail to make a full, understandable sentence here.
- '...Some of them also with an Aurovist...' on line 16 should state *what* has been done with Aurovist/Omnipaque; instilled, washed, show to?
- 'enhancing soft tissue structures' on line 19 should probably state 'enhancing the visualization of soft tissue structures'
- 'that there is not' on line 20 should probably be something like 'that it is not'
- 'very significant' on line 21 should only be said so if you can state a p-value of significance, especially in the abstract.
- 'it seems that the use of ... could increase' on line 23/24 need *much* more substantiation than this, especially in the abstract. The authors should mention what the success rate is.
- Why are the chicken embryos not mentioned in detail the abstract, but only in passing of line 15? The details of those scans are 9 of 13 tables in the appendix...
- The fuzzy logic approach should *not* be mentioned in the abstract; I was waiting for this while reading the whole manuscript, and it only shows up in the paragraph about 'future research', which is misleading the reader of the manuscript.

### Introduction
- Line 41: '...is in many cases is not...' suffers from a duplicate 'is'. The whole sentence is thrown onto the reader, noting about the noise in the detectors is explained more.
- Line 43: What are 'CT numbers'?
- Line 48: What exactly does make the segmentation process easier here? The *whole* manuscript is only about visualization, segmentation is a whole different process. It seems that the authors haven't understood the underlying issues properly, mixing up segmentation with visualization.
- The last paragraph starting on line 51 is mixing transfer functions with LUTs (which kinda are the same thing) and the idea with the glass capillaries, which is in itself not really relevant for the main topic of this manuscript.

### Materials and Methods
- Is the database mentioned on line 59 publicly accessible?
- Line 65; As it is the first time the SkyScan 1275 is mentioned, a bit more details (supplier, etc.) should be mentioned. I would also not mention it as 'kit', but as 'system' or something else.
- Line 69: The details of the intravenous instillation *have* to be mentioned, as they are quite crucial in how the contrast is then shown in the tomographic data.
- Line 76: What are the 'slides' that are mentioned here? Tissue sections, or light microscopy slides? Needs to be made more precise.
- Line 82: Figure 8 should not be the first figure that is cited in the text. Make this figure 1.
- Who is supplying the Phywe system (line 83)? The reader needs more details about this system.
- Line 85: I guess *all* systems by Bruker are for professionals, not hobby-tomographers. Additionally, I think that the denominator of the Bruker system (e.g. 1275) is missing.
- Line 91 should be 'The anode power range is 20-100 kV at 10 W.'
- Line 94: it is instead of it's
- The sentence from Line 95 to 98 is very long and very hard to understand.
- Line 109: '0-1 mA' would probably be better stated as 'up to 1 mA'.
- Line 112: '...and transfer function' seems to be missing a verb.
- Line 128: No abbreviations in titles.
- The first paragraph of section 2.1.1 instills a sense of dumbness onto the reader. Certain things can be assumed.
- The equation on line 139 should not have a '.' as multiplication sign.
- Line 141: 'how steep line is' should be 'how steep the line is'
- The equation on line 142 should not have an unmotivated space.
- The text between line 150 and 160 feels like it would show up in an undergraduate textbook and should *not* be in a scientific manuscript.
- Line 170 is one of the main gripes of mine with this manuscript: 'segmentation' is *not* the same as 'visualization'. By influencing the LUT you are only changing the representation on screen and not doing and proper segmentation. The process of segmentation can be a binarization or labeling of the data, but has nothing to do on how it is represented on screen by a LUT.
- Line 174: 'Figure 3' does *not* contain a panel 12; maybe the authors want to mention Fig. 4?
- Line 188: The human eye can distinguish *much* more than 256 gray values, a cursory internet search gives numbers > 100000: https://hypertextbook.com/facts/2006/JenniferLeong.shtml. Can you insert either a citation here or tell the reader why you've used a mapping to 8 bit and not 16 bit?
- Line 196: 'NRecon SW software' is repeating the word software unnecessarily.
- Line 210 and 210: 'each DICOM browser' and 'any DICOM browser' are unnecessarily repeated.
- The whole text from line 220 to 235 is *only* valid for CTvox, other software has (completely) different ways of influencing the LUT. Either generalize or state that this is the case for *only* CTvox.

### Results
- Why did you design 53 transfer functions, not 55 or 33 or 111?
- Why did you design 11 functions that are useless for your samples?
- Line 244: 'one TF was able to be suggested that would likely highlight other interesting' is a sentence that can hardly be made any more vague. The authors should state what they did and why and not suggest things that by a random chance could possibly highlight other things!
- Figure 8: It's nearly impossible to see all the capillaries in the top image. They should probably be marked. The figure is missing a scale bar or (better) a scale reference. There is much too much white space on the left and right side of the image. What's the difference between the green silicone and the transparent silicone?
- Figure 10: There's *no* way that any reader can gain *any* meaningful information in the right side of the figure. The difference in the subpanels 1 and 2 has to be explained in the figure caption.
- Figure 11: What is a 'mouse hull' in panel 2? Did the authors mean a 'slice/region through the thorax of a mouse'?
- Figure 12: There is *no* information of the influence of the difference in acquisition voltage on the visualization, the results are thrown towards the reader. The blue/red contrast is *very* misleading in showing differences. Any perceived differences might be simply due to the different LUT *or* due to the different acquisition voltage. Much more has to be explained here.
- Line 294: Where is this segmentation of the capillaries inside the chicken flesh shown? This is again not a consequence of the LUT, but simply of the sample and scanning parameters. If the resolution is good enough and the attenuation contrast in the sample is high enough, then the capillaries can be detected/segmented, it's not because of the LUT!
- Line 296: 'blood vessel' should be 'blood vessels'.
- Line 297: 'could increase the success rate' is - again - very vague. How and why is this better?
- Lines 303-309 are simply repeated from earlier. Honestly, it seems that *none* of the authors did a proper proof-reading...
- Lines 310-312 are simply repeated from lines 294-296. Again, did *someone* proofread this before submission?
- Lines 321ff: Having huge amounts of data does not automatically make segmentation possible. The conclusion in this paragraph is bogus. Why is the part with 'staining of the head scan of a chick' even included if the authors don't know what they saw? And 'staining of the head scan' is wrong, it should say 'scan of a stained chicken head' or something like this.
- Lines 354-355 should be removed from the manuscript, why is it even there when nothing is to be said of those samples?
- Why is the idea about 'fuzzy approach' also in the abstract if it's only a hint on further research to be done?

### Conclusions
- Why do the authors mention ultrasound contrast agents? This is the first time the authors mention this and it has nothing to do with the conclusion on the data from their scans?
- Where is the frog's heart coming from on line 371? This is the first time that a frog is mentioned in this manuscript! Again, did *someone* read the full manuscript?

## Data Availability Statement:
The three sentences in the data availability statement are contradicting each other.
It seems that the authors just copied them from a template and failed to remove two irrelevant sentences.

## Appendix
- Figures A4, A5, A6 and A7 are only different LUTs, no? Why are they added to the appendix without an explanation of their LUTs? In all these figures Aurovist is not hyphenated but printed on two lines.
- Figure A9; Why do the two lung lobes have a different density? Is this because of the Omnipaque instillation or because of some disease in the animal?
- Figure A10: Again, this is *not* a segmentation, only a visualization.
- Figures A10 and A11 (and others): The font size of the scale bar label is different. It would be nice if this is homogenized throughout the manuscript.
- Figures A12, A13 and A14: What's the difference between the left-right or top-bottom images. Why are *these* images shown and why is there no explanation/discussion on the differences. 

# My conclusion
This manuscript seems like too hastily put together and tries (but fails) to tell an interesting story hidden in the data acquired by the authors.
I'm not even sure that the authors *did* acquire all the data or one person fiddled with lots of LUTs and applied them to tomographic scans pulled from a library.
Several very major issues with the manuscript fail to convince me that this is worthy of publication in a scientific journal; certain ideas could be fleshed out and be added to a learning resource about *visualization* data (and not segmentation of data).
I think that I might be able to generate LUTs that show the data exactly in the same way for both the Aurovist- and Omnipaque-instilled samples, even though their contrast is different. Again, I fail to see the scientific advance of looking at 53 LUTs that 'fell from the sky' and reporting 'those look good'. 

Author Response

Please see the attachement.

Author Response File: Author Response.docx

Reviewer 2 Report

This paper presents a study to optimize segmentation and soft tissue contrast visualization of biological specimens. 

 

The following comments and concerns where ever applicable shall be considered to improve the quality of the paper. 

• The overall reading of this article seem to require more attention and a thorough language editing assistance. There exists statements that mislead readers. For instance, Please check the language and grammar context in line 13-14, 51-52 and so on.
• I would suggest authors rewrite abstract with a flow. 
• Introduction section seem to have missing the narrative. Although the problem statement seem to be clear (line 39-44). However, i would suggest authors to list the clear contributions of the proposed work in this article. Add more on how rate of segmentation is better achieved through the proposed work. 
• Related study seem to be missing. I would suggest authors to include a related study section and highlight the similar work carried out in the past to emphasize the need of the proposed work. 
• I would suggest authors to cite a link (line 59) to access the database used in this study for better reproducibility of the intended study.  Details regarding the devices and software used seem to be fine. 
• I would suggest authors to reconsider to employ the images used specifically Figure 2, 3, 4, 5, 6 and 7 effectively and mention there usage appropriately. 
• Also, Please check figure usage references carefully. For instance, Figure 9 usage reference seems to be incorrect. Please Check line 227, and 235. 
• The narrative used to present transfer functions, LUT concept seem to be good. 
• I would suggest authors to summarize/tabulate  the findings and explain the advantages of the proposed method used in this study.  Please consider to reorganize figures. 
• Discussion section seem to be fine and highlights the limitations well. 

 

Author Response

Please see attached file.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

## General remarks
Thanks for working on the manuscript based on my suggestions for `v1`.
Unfortunately, large parts of the manuscript still feel to me like they would be more at home in an undergraduate textbook and - in my opinion -fail to highlight new scientific results apart from a thorough description of experiments.
I don't see anything that I could take for my work (also in microtomography of biomedical samples) apart from 'one should experiment with lots of LUTs'.

## General issues
Even thought 8 (!) authors claim to 'have read and agreed to the published version of the manuscript' it would profit from some *very* thorough proof-reading and rework.
In at least one instance, there's a repeated sentence in one paragraph, crucial preparation steps for the biological specimens are omitted, details about the tomographic data acquisition are repeated in needless detail and a large, overarching story is missing from the manuscript.

I fail to see any scientific contribution in this manuscript and would only want to read it again after the authors have fixes *several major* concerns of mine.

## Detailed comments
### Abstract
- Thanks for completely reworking the abstract.
- Why is the PHYWE system not mentioned?

### Introduction
- No further comments

### Materials and Methods
- Thanks for providing the LUTs, but they do not help in reviewing this paper. You could think about providing them to the reader of the paper.
- I'm quite well aware of how contrast agents work.
- You mention that 'Aurovist is *inhaled* into the tail vein'. Do you mean *injected*, or is Aurovist going into the lung?
- The sentences "Subsequently, scanning is performed due to the contrast of the gold nanoparticles that are contained in Aurovist. The substance is therefore dissolved in the blood and so both arteries and veins are visible." do not make sense. 'Scanning' is not 'performed due to the contrast', it is due to the gold nanoparticles in the Aurovist in the blood stream (probably *not* dissolved) that the blood vessels show up under microCT imaging.
- I think that for both systems (Bruker and Phywe) you should mention exactly where they come from. Usually in a manuscript, I would expect to read "SkyScan 1272 (Bruker MicroCT, Kontich, Belgium)" or "Phywe XR 4.0 (Phywe, "Some City", Germany)".
- ORS Dragonfly is mentioned, but no statistical analysis or LUTs from this software seem to be done/presented. Is that true?

### Results/Discussion
- The statistical analysis seems very 'bolted on' and needs more reasoning about hypothesis and testing.
- I would *not* expect any influence of the Ethanol concentration onto the visualization. You confirm this, but should mention this at least as hypothesis to be tested (if that's one of the aims of the manuscript).
- Lines 379ff: Is the failure of visualizing soft tissues or organs really related to the size of the data? Or is there some issue with sample preparation and/or scanning? Data larger than this *has* been visualized...

Author Response

Thank you for your valuable notes on the manuscript. We tried to reply to all of them in attached file.

Author Response File: Author Response.docx

Reviewer 2 Report

I value authors efforts in revising the article considering reviewers comments/concerns. The overall article now exhibits better readability and authors have addressed my comments well. Table 1, 2, 3, and 4 add comparable aspects presented in this article.  I would suggest authors to use high quality images wherever possible (Ex: Figure 9). I have no further comments. 

Author Response

Thank you for your valuable comments on the manuscript. We tried to cover it and update manuscript as best.

Author Response File: Author Response.docx

Round 3

Reviewer 1 Report

Thanks for your further work on the manuscript.

Why are you 'suddenly' mentioning ultrasound contrast agents and fuzzy logic again (around lines 65ff) when it's the only time that is mentioned in the manuscript? 

Lines 77ff: Why is H2 not mentioned with 'H2', but only in text?

Line 136: What was the concentration of potassium iodide in the final solution? This is quite crucial for assessing the contrast.

Line 380: Authors should not speak of 'I' in a scientific manuscript, especially when there are several authors. It seems that this paragraph was simply copy-pasted from an email. The URLs could be replaced by proper citations.

Line 489: "and it's the limitation as well" is rather colloquially written. Which two concentrations were used?

Author Response

Dear Reviewer, attached please find step by step responses to your comments.

Author Response File: Author Response.docx