Influence of Uncertainties in Optode Positions on Self-Calibrating or Dual-Slope Diffuse Optical Measurements

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This work focuses on a quantitative analysis of the impact of errors in source and detector positions on the assessment of optical properties of highly scattering media and biological tissue. I think it is interesting and can guide the optimal design of optical probes for self-calibrating and dual-slope measurements. However, there are There are still certain aspects of this paper that require further refinement.

1. The full text relies on theoretical simulations (based on diffusion theory), However, incorporating experimental data would enhance the credibility of the results. For example, does a 1 mm displacement truly cause a 19% error in the scattering coefficient? Can this conclusion be verified experimentally?
2. In Tables 1–6, the simulated ranges of μₐ=0.005–015 mm⁻¹ and μs'=0.5–1.5 mm⁻¹ do not cover high-absorption tissues (such as vascular-rich areas, where μₐ can reach 0.02 mm⁻¹) or low-scattering media (such as cerebrospinal fluid). Is it possible to expand the parameter range or cite literature to explain the matching of the selected parameters with actual biological tissues.

Comments for author File: Comments.pdf

Author Response

See attached pdf

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors,

the manuscript under review presents some analysis of alignment sensetivities for an optical system that uses two optode pairs to determine scattering and absorbtion of a biological sample. In general, this topic matches the scope of Photonics journal, but there are a few points, which should be revised or clarified before the paper could be accepted for publication.

MAJOR

1. First of all, I have some doubts about the model. The sources and detectors are represented as perfect points, but actual elements will have finite sizes, and , that is more important,  direction diagrams. So if one consider a more realistic representation of, let's say, LED sources and  photodiodes, the model will be also sensitive to angular positioning of those. I suspect, that the simplified model is not representative and leads to a wrong errors budget.
2. The approach to analyze the sensitivities also leaves some questions. If the model is simplified and there are just a few parameters describing it, could it be possible to find teh sensitivities to misalignment analytically or semi-analyrically through derivatives (see section 2.3.1)?
3. Simulataneous influence of multiple errors is what will happen in the reality, so the results in Table 6 are more important than all teh preceeding calculations. However, in practice such an analysis is performed through a Maonte-Carlo simulation or a root-square sum of errors, but not by manual probing of different errors combination.
4. Notably, the errors have no statistics. Normally, one could expect that each positioning error is a random value with some statistical sdistribution (e.g. Gaussian or uniform). Instead, these variables atre treated as detecrmonistic ones.
5. As a consequesce of the previous points - for me  all teh tables and plots in Section 3 are just intermediate results, necessary to understand the sensetivities of the measurement system, but not enough to make firm conclusions on the precision requirements. In imaging and illumination systems design, such an analysis normally represents teh first stage of the system tolerancing. It is a relatively trivial engineering task, which results in some requirements for manufacturing, assembly and alignment. At the moment it seems to me that the analysis is not comprehensuve enough and has relatively low value for teh readers.
6. In turn, as too high attention is paid to the preliminary sensitivity analysis, the paper becomes too long. I have some doubts that the tables and plots in Sec.3 showing individual sensitivities would be of a high value for the readers. In particular, they seem to be quite dependant on the nominal geometry of the optodes system. I would strongly recommend to rework this part and produce fewer summary plots or tables showing the critical sensitivities depending on the geometry and analysis method.

MINOR

1. It seems that the measurement method is applied mainly for biological samples, but it is not refected in the title and keywords.
2. Fig.1.- it would be very useful to show the sample position in the schematic drawings.
3. Line 218 - it is stated that the iterative method has a priority, but further bothof them are tested in details.
4. Fig.10 and 11 and teh corresponding text - it sounds as if there was a goal to buold a scanning system and find some equidistant trajectories. Would it be easier to use just a very rigid common frame with no preferred directions?

To summarize, I see that the model has some serious flaws and the errors budget can be incomplete and incorrectly analyzed. An intermediatee stage of the tolerance analysis is over-presented, while the paper still misses some clear results valuable for practice. Therefore I cannot recommend the current version for publication and advise the Authors to rewise their design and analysis, correct the manuscript accordingly and resubmit it.

Author Response

See attached pdf

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

In the manuscript titled "Influence of Uncertainties in Optode Positions on Self-Calibrating or Dual-Slope Diffuse Optical Measurements", the authors describe a simulation study to evaluate the influence of optode positioning on the assessment of optical properties. The work is likely of interest in the field, as self-calibrating and dual-slope methods provide calibration free measurements of optical properties, and a better understanding of potential systematic errors in measurements are important to understand.

While the results presented are interesting, before recommending this work be published, I believe it would be greatly beneficial to see the results presented in the context of typical instrumentation noise, especially given the relatively extended source-detector separations explored in the work, to better understand the order of magnitude in the errors relative to what would be expected given the noise of a typical instrument. I think without this comparison, it is difficult to evaluate the relative error introduced by assuming incorrect distances. 

In addition to understanding the effects of instrument noise on the estimates of optical properties, I have a few minor comments to address as well.

1. For the figures demonstrating the error in the estimation of optical properties with respect to the 2 cm x 2 cm grid, would it be reasonable to to add subplots that compare the average error in source-detector separation and the error in optical properties? The heat maps are very illustrative for demonstrating the concept of the iso-lines, but I think it would also be nice to see if particular displacement directions are important because displacements along those directions cause greater changes in source-detector separation or if there is something more fundamental related to the geometry of the probe that results in the importance of those displacements.

2. In the discussion, it would be nice to have an expanded discussion on the effect of tissue curvature on both the positioning of optodes as well as their pointing direction relative to each other and the relation of these concepts to the presented simulations. 

 

Author Response

See attached pdf

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors,

 

Thank you for sending your response to the first round comments.

As far as I can see, the changes are rather minor, so I suppose, I should explain some of my points from the first review.

1. I agree that the actual influence of the real source and detector propertiess may be small and even negligible. However, such a statement requires some proofs and numerical estimations. The effect of finite sizes was analyzed and briefly covered in the revised manuscript. However, the aspect of angular alignment is still ignored. I've tried to make a quick estimation using a raytracig model with a 0.3 mm source, a perfect plane with scattering properties following the Lambertian law and a 1.5 mm detector. The detector is shifted from the source by 20 mm in Y direction, the Z distance between their common plane and the sample is subsequently set to 25,50 and 100 mm. I've used the direction diagram of a commercial Avago HLMP-HG64 LED for this test, since there was no better reference. Indeed, the sensitivity to the source tilt is quitesmall , although it is not zero. The results are as follows: at 25 mm distance a large 10 degree tilt leads to only 2.3% change in the total irradiation at the detector; at 50mm it becomes 2.1%, at 100 mm - 3.5%. At the same time the absolute values change as well - by factor of x0.52 when moving the target from 25 to 55 mm and by another x0.59 when moving from 50 to 100 mm. This simple test reveals a couple of important points

1.1 The actual direction diagram can have different steepness and even be anamorphic (as the one in this example). The actual distribution may have some influence on the output

1.2 The source-to-sample distance has a major impact and it is still not provided in Fig.1. and the following text. 

1.3 The angular alignment influence is small, but still should be mentioned in the text

2. The rest of the discussion basically reduces to the reply #2.3. It seems that the approach to errors analysis in this case is drastically different from that used for some other optical systems. In partiular, I would like to focus on the following statement: "random errors are not the focus of this work". So, the question is - what kind of errors are in the focus? I can assume a case when the system is pre-aligned in a laboratory and then used only on one specific sample. Then, indeed, the errors will be deterministic. If the system is designed for use on multiple samples, or even multiple regions of the sample (as I conclude from another reply), these installation misalignments will  satisfy the definition of a random error. And all of them (dx,dy,dz at each elements as well as tilts) will appear simultaneously. 

One may think about other non random errors as a thermal drift. But these one can be taken into account analytically at the design stage and/or calibrated during the operation. 

Until the nature of the errors and their modelling are clarified, it may be difficult to discuss on comments 2.4-2.10.

3. I still tend to consider the results obtained here as intermediate ones. It remains doubtful that they can be useful for a reader. In particular, the first of the results mentioned in the conclusions, i.e. "estimation of the impact of uncertainties in the optode positions on absolute or relative optical measurements", provides only general qualitative dependences as the general advantage of iterative method. Any quantitative  results will be inapplicable for another system with a different geometry, real components and a different sample. So, one will have to re-do a similar analysis. 

The second result is "guidance for the design of optical probes to minimize probe deformation along directions". This result could be of a higher practical significance, but, unfortunately, the design proposal does not go further than a basic sketch.  Perhaps, if the low sensitivity to misalignments caould be demonstrated in a simple prototype using this geometry, it could be of a higher interest to the readers. 

Since the changes in manuscript are rather minor I can't change my overall opinion. However, I believe that adding even a short experimental part could make the paper suitable for publication regardless of some simplifications in the modelling. If the authors accept this point of view and the corresponding revision plan, I would also strongly recommend to shorten the existing numerical analysis to make it easier to read. 

 

 

 

 

Author Response

See attached

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

I'd like to thank the authors for their attention and effort to my comments. To me the work is appropriate for publication.

Author Response

See attached

Author Response File: Author Response.pdf