Development and Evaluation of a Dual-Layer-Offset PET Detector Constructed with Different Reflectors

Round 1

Reviewer 1 Report

Dear Authors,

 

many thanks for this nice paper. You can find few comments below.

 

Line 13    mm3--->please use the uppercase for the power

Line 16    center of gravity of the crystals? threshold of the the analog waveform? to generate time and enery histogram? please explain better

Line 17    crystal identification is not a  performance, it is parte of the optimisation.

Line 37    William W. Moses, please cite his paper https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3144741/

Line 57     CTR if you write CRT this is Coincidence resolving time

Line 108    crystal strips: why strips? isn't crystals enough to refer to them? In my mind strips is a well defined gemetrical configuration. Your crytsals look more like fingers than strips

Line 140    by air: why by air and not with optical grease or melt mount?

 or something else?

Figure 3    Please name the axis and include units of measurements where possible.  Please refer to the image a and b in the caption. This comment is also valid for alla the figures. I'll not accept a paper without name on axes of the plots. 

Line 174   I think I have understood, but I would like to read more details about the extraction of this LUT which is a crucial point of your work.

Figure 4    please center images with the text. This comment is valide for all the figures

Energy resolution: what about ESR, even if it's bad it's nice to compare to the others.

 

conclusions and discussions: 

Very nice. Anyways why don't you provide an estimation of the DOI and of the source position precision in x and y? For the first one it would be better to have a simulation, even if It would be nice to see at least a distribution of the measured DOI.

For the second one, it is true that the precision is determined by the pitch of the crystal, but it would be nice to see the effect of the double layer.

 

 

 

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

This manuscript compares 3 reflector types on dual layered offset PET scintillator arrays. The paper is concise, easy to read and the English is ok. However, I have several concerns, including a few major ones.

1. My main concern is regarding the novelty or advancement of this manuscript. Offset crystals for depth of interaction information in PET detectors has been widely studied since first proposed by Chris Thompson et al., And so has the reflectors listed by the authors. Furthermore, a recent paper reported by Kang et. al., (“A staggered 3-layer DOI PET detector using BaSO4 reflector for enhanced crystal identification and inter-crystal scattering event discrimination capability”, Biomedical Physics & Engineering Express, 2021) reported a very similar study. In fact this paper reported a three layered configuration, and compared the performance of similarly size crystals and similar reflectors (a specular ESR and the diffused barium sulphate). The only difference is the addition of the lumirror reflector but that too has been widely studied. Please discuss what is the novelty of this manuscript.
2. Why isn’t a light guide used between the scintillator and SiPM? It would seem that a light guide would help light spread across the SiPMs and improve positioning. This is especially important for specular reflectors such as the ESR where the light spread is limited, compard to diffused reflectors such as barium sulphate. The same applies for Lumirror reflector which is in-between a specular and Lambertian reflector. Maybe that is why the positioning of ESR (and lumirror) reflector array was so poor. Did the authors test using a light guide? If so, please show results. If not, why?
3. In the paper by Kang et al above, the energy resolution with ESR reflector was excellent and individual elements in the flood map of a 3 layered offset detector could be identified. What do authors think is the difference between this study and that paper that lead to the different outcome?
4. Why is the energy resolution of lumirror reflector scintillators so poor? I would expect that with its high reflectivity, the energy resolution would be better than BaSo4. Could is due to the manufacturing process of the crystal arrays?
5. In PET, the timing resolution, in addition to the spatial resolution is also crucial. However, it is known that ESR reflector typically provides better timing resolution than BaSo4 reflectors. Did authors compare the timing resolution between the blocks? Since there are likely tradeoffs (energy/spatial resolution vs timing resolution) involved, I recommend that the timing resolution also be provided, especially in a study like this which is a straightforward comparison of reflectors.
6. Lines 46-53 is an exact copy from reference 9. Authors must rephrase the sentences.
7. The journal Crystal focuses on “the advancement of our understanding of the nucleation, growth, processing, and characterization of crystalline and liquid crystalline materials”. This manuscript however, is on the comparison of reflector types, using similar LYSO scintillation arrays. The performances of scintillation crystals are reported but the main target of interest here, is the effect of reflector on the scintillator. Similar papers have been published in Crystals but are usually part of a special issue. What is the authors opinion on this?

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Review for the manuscript:

Entitled: "Development and evaluation of a dual-layer-offset PET detector

constructed with different reflectors"

 

for Crystals.

 

With ID: crystals-1507550

 

Dear Authors

 

Thank you for your manuscript.

 

General comments

 

Comments for the Authors,

 

This work is well within the scope of Crystals, and it may be of interest to most of the readers of this journal. It shows an introductory background material sufficient for someone not an expert in this area to understand the context and significance of this work, with good references to follow. Turnitin showed a similarity index of 32%, especially in the introduction and materials sections.

Furthermore, the introduction section lucks to justify the choice of LYSO for the examined system.

My main concern regarding the publication of this study is its aim.

Authors state that: ‘This paper provides referable guidance for choosing scintillator reflectors for multi-layer dedicated DOI detectors designed for small-animal PET imaging.’ and the study concludes that: ‘the results indicated that the BaSO₄ reflector performed substantially better than the other two reflectors in both crystal identification and energy resolution’ However it is well known that BaSO₄ performs better than other reflectors, as it was also stated in a study published this year:

 

‘A staggered 3-layer DOI PET detector using BaSO₄ reflector for enhanced crystal identification and inter-crystal scattering event discrimination capability’

https://iopscience.iop.org/article/10.1088/2057-1976/abf6a8

Authors should address these issues and the specific comment below, prior the publication of this study. Thus, I have opted to recommend a Major Revision for this manuscript.

 

Specific comments

 

P2, L46-49: ‘In a system with a certain depth of interaction (DOI) resolution w, the radial resolution at any given point within the field of view can be estimated as follows, assuming perfect reconstruction and individual coupling and readout and neglecting the impact of positron range and the angular nonlinearity of annihilation photons’ Please revise this paragraph.

 

P5, Authors state that ESR reflectors showed poor decoding performance, however it could be of worth to provide Peak-to-valley ratios, since they are provided for the other two reflectors.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Dear Author,

Thanks for your reply. Below few comments.

Point 2: I would say "An optimized interaction-decoding algorithm using the center of gravity to determine the position and threshold of analog signal for timing methods was applied to generate decoding flood histograms "

Point 7: If properly done it is not true that measurements  with optical grease have reproducibility problems

Point 8: The fact that most of the people present it like that does not mean that it is correct

Point 10 Some figures are not yet centered.

Point 11: If you include the spectrum for ESR, do it for other configurations too. Moreover this bad result is linked to the fact that you are not using a coupling medium. In general I understand what you want to show and that the comparison between wrapping technique is relative and not absolute. 

 

Author Response

Please see the attachment

Reviewer 2 Report

Authors have addressed many of my questions. There are a few more issues left, especially that of the novelty of the manuscript and similarity compared to an earlier paper by kang et al., (which the authors still did not reference), the following are my questions:

 

Novelty
1. Authors replied that the novelty of the paper lies with the use of the thresholding method. However, if this method has been reported in  ref 44, it is no longer new (in this paper at least).

2. Was no coupling agent or light guide whatsoever used between the crystal layers and also in the crystal-sipm boundary? Please clarify.

If so, this could simplify manufacturing but at the cost of performance as evident compared to the paper by Kang et al (“A staggered 3-layer DOI PET detector using BaSO4  reflector for enhanced crystal identification and inter-crystal scattering event discrimination capability”, Biomedical Physics & Engineering Express, 2021).

Thus the above paper must be referenced due to its similarities, and the differences (no optical glue or light guide used; three different types of reflector compared) stressed.

 

Threshold value:

Are the threshold values different for each SiPM pixel? It seems like a fixed value is applied for all sipms. In this case, why not simply add up the signal and then apply the threshold to that value to get rid of small noise signals?

 

Timing resolution:

It is surprising that Baso4 led to better timing resolution than ESR scintillators. I do not recall seeing papers reporting better timing resolution with Baso4 for such high aspect ratio scintillators (long and thin crystals). What could be leading to this? Authors suggest “the absence of the light guide and the optical grease between two layers of crystals lengthened the scintillating photons’ paths for ESR reflectors”. However, I think otherwise. Can authors provide reference supporting this claim and that timing resolution is poorer with ESR (compared to Baso4) with similar aspect ratio scintillators?

 

Flood map:

In figure 3, the counts in the top layer is missing. Also, what the p/v graphs correspond to? The columns or the rows?

 

(minor) English:

Grammatical and typographical errors are still visible.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Review for the manuscript:

Entitled: "Development and evaluation of a dual-layer-offset PET detector

constructed with different reflectors"

 

for Crystals.

 

With ID: crystals-1507550.R1

 

Dear Authors,

 

General comments

 

Comments for the Authors

 

My previous comments were addressed; thus, I have opted to recommend the acceptance of the manuscript

Author Response

Thanks for the reviewer's comments.