High-Sensitivity X-ray Phase Imaging System Based on a Hartmann Wavefront Sensor
, Martin Piponnier 3,4, Laura Oudjedi 3,4, Xavier Levecq 3,4, Fabrice Harms 3,4, Alessia Cedola 2, Ombeline de La Rochefoucauld 3,4 and Philippe Zeitoun 1,*
Round 1
Reviewer 1 Report
This paper well demonstrates an advantage of a Hartmann wavefront sensor. The author focuses on an effect of the reference images and clearly shows the impact of them. However some explanation is not sufficient and there is still room for improvement.
The followings are comments and questions.
Through the whole paper:
There should be an indentation at the begin of a paragraph.
2.Material and Method:
Readers may like to know more detail about the experimental set up, such as information of the X-ray source, (its spot size and wave length etc), and of the Hartmann Sensor (the number of the holes etc.).
- Results
3.1 Impact of reference image:
The setup for the measurement in water is not clear. I suspect that the grape is in a Ependorff filled with water and that other region except for the inside of the Ependorff is in air. Is this correct? Then how did the authors measure the reference in water? Does “the same cylindrical container” mean the Ependorff filled with water? Anyway readers need more clear description.
The difference between with and without median filter is so large between b1 and b2 in figure 2 (and also figure 3). The authors should mention an explanation of such difference.
The image in b2 of Figure 2 shows two dark spots around a little below the right. What are they?
The authors mentioned about moire pattern, but it is difficult to recognize the pattern in the figure.
From line 114 to line 118, the authors are discussing about the moire pattern, but fig.3c2 (mentioned in line 115) had been applied with the median filter, and in line 117, fig2 suddenly appears. Can readers see a moire patter in fig. 2 ? Also I think that the effect of the median filter onto the moire pattern may depend on the size for the calculation of the median. Do the authors need to discuss about the size of the median filter? Anyway this paragraph is very difficult to understand the intention of the authors.
3.2 3D Imaging
The setup picture is not clear. Is there a snail on the table shown in the setup picture? Where is the Hartmann Sensor and the X-ray source? Please explain more.
In line 132, the authors mention “kernel area 200 pixels”. Is this the same to the “size 200*200” as wrote in previously? If they are the same, please unify the description and if not, please explain more.
From line 135 to line 137, the authors mention about the dark fringe and moire fringes, But readers cannot recognize them.
- Discussion
From line 167 to line 178, the authors discuss about the moire pattern, but I cannot catch the point what the authors like to intend. The moire pattern is trivial phenomenon and the description in this paragraph does not contain any new information. Do authors discuss about the demerit of the Hartmann sensor? A Hartmann sensor with a pixel detector like a CCD should have this moire problem, more or less.
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Questions/clarifications It is not explained how it is possible to exploit absorption, together with deflection method. It is measured the different intensity of each deflected beamlet? The authors can easily clarify in the text. Line 56 Can you specify the properties of the detector? (size, spatial resolution). If written here the following would be more clear (see also below other related questions) Line 58 It is often used the word "sensor" to indicate the Hartmann array, here sensor indicates the Harman plate or the detector? If this is the size of the Hartmann plate, what is the size of the detector? Line 88: I understand Test1 and Test3 in which you take the reference image the same as the sample, but why to take a reference image in different conditions with respect to the sample measure (Test 2)? What is expected to improve if the medium is different? The authors can explain in the text why they carry out measurement in these different conditions. Line 106 - You measured similar deflection in X and Y, therefore the images shown in fig 3 are only for X. Is there any improvement in the image reconstruction if you use the distance deltaZ = sqrt(deltax^2 + deltay^2) instead of its projection along x or y? I mean, I expected that more contrast is obtained if I consider the net distance between reference and measure, not the projections. Line 124: What it means the spatial resolution on the sample plane is 37.5 um? Is 37.5 um the spatial distribution of the holes in the Hartmann mask? If yes, I suggest to write it at the beginning of the document, for instance around lines 58-59, instead of telling through the text (it is told twice) Line 127 - Fig. 4 Not so important, but I am not sure I see the sample in the small red box. More importantly, the setup would be more clear if you put labels in the picture just to indicate the flat panel, the Harmann plate ecc...I recommend the following publications to be cited for x-ray imaging with Hartmann plates:
Mikhaylov et al. (2020) https://doi.org/10.1107/S1600577520002830
Wilkins et al. (2014) https://doi.org/10.1098/rsta.2013.0021
Style corrections General comment: I would replace the words fig.2, fig.3 with Fig. 2, Fig. 3 (upper case and space between Fig and number. Line 16: ... with an array... Line 36: ...take advantage of ... (or ...imaging benefits from...) Line 39 loosing --> losing Line 44: this kind of sensor... Line 59 no need to specify "Impact of reference image". Sect. 3.1 is sufficient. LIne 61: no need to specify "3d imaging". Sect. 3.2 is sufficient. Line 84: 37.5 um instead of 37,5 um. Line 85: I did not know what an Ependorff was, then I checked online and I found that is a commercial name for tubes. Just use "tube". Or, if you prefer, using the correct name "Eppendorf tube". Line 90: --> these experimental conditions Line 98: ... structures are hard... Line 105 and 107: axis --> axes Line 112: no need to specify section 4 'Discussion'. Section 4 is sufficient. LIne 121: .. on a test sample ... Line 147 - Figure 5 caption: "...wavefront sensor. The..." (space between sensor and The) Line 156 ... needs to be done Line 158: ...for a long time. Line 173: bended --> bent Line 184: ... measurement needs ... Line 189: ... multiple sources...
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
The authour revised the manuscript according to my comments. Although the revision helped to understand their analysis on the moire fringes for me, the revision complicated the flow of description. The author should revise again in order to make the description to be simple flow.
- Please add an indent at the beginning of each paragraph.
- L60-61
The sentence - “The spot size is 20 µm and the emitted bremsstrahlung spectrum is characterized by the Kα of Gallium at 9.2 keV”- is wrong. I guess “bremsstrahling including the Kα of Gallium at 9.2 keV”
3.L103-104
If the cylindrical container is the same with the Ependorff tube, this sentence is a repetition of the description in L96-L97.
- L105 and later
The order of the figure should be changed. I think fig 4 should be moved prior to fig2.
- L116
The term, “moire fringes”, suddenly appears. But readers see the moire fringes first in figure 2 a. Thus please mention the moire fringe in the explanation of figure 2a.
By the way, why does the image in fig 3 (left) differ from that in fig 2 (a)?
- L121-L127
The description flow is complicated because it goes back and forth. Please reconsider the description flow.
- L 133-L136
The author mentions the moire patterns again, as it appears in the first time. This should be revised.
- L160
Authors again mention the moire fringes, but readers can not recognized the moire fringes in the figure 7a.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
