Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Understanding the distribution of water channels in the cornea may provide useful insights into ocular adaptation to arid environments. The manuscript investigates the spatial distribution of aquaporin-1 (AQP1) in different regions of the dromedary camel cornea and the topic is therefore potentially interesting for comparative ophthalmic histology and veterinary anatomy. However, in its present form the manuscript contains several methodological, analytical, and presentation issues that significantly limit reproducibility and interpretation of the results. In particular, important details of the immunohistochemical protocol, quantification of AQP1 expression, and the claimed use of “AI tools” are insufficiently described. Furthermore, several problems are present in the preparation and annotation of histological figures.

Substantial revision is therefore required before the manuscript can be considered for publication.

The manuscript states that corneas were obtained from “healthy animals,” but no criteria are provided describing how the health status was determined. It would be important to clarify whether animals underwent veterinary or ophthalmological examination prior to slaughter and to provide basic information about the animals (for example: age, sex, and general condition). In addition, the post-mortem interval and the time between tissue collection and fixation should be specified. Although the manuscript states that the study was approved by the institutional ethics committee of Sadat University, the ethical approval number or protocol identification is not provided. The official approval number and the date of approval issued by the ethics committee should therefore be included.

The cornea was divided into nine regions (central, middle, peripheral; dorsal, ventral, nasal, temporal). However, the manuscript does not explain how these regions were anatomically defined. It would be useful to clarify several methodological aspects: how the boundaries between regions were determined, whether a standardized template or measurement grid was used, and how the orientation of the cornea was preserved during sampling. Without these details it is difficult to evaluate the reproducibility of the regional comparisons.

Important technical details are also missing from the description of histological processing and morphometric analysis. The manuscript should provide additional information about the histological methodology (for example: the type and manufacturer of the microtome used for sectioning, the criteria used to exclude oblique sections, the number of sections analyzed per animal, and the number of microscopic fields evaluated in each region). Without these details the morphometric measurements cannot be reproduced.

The immunohistochemical method is insufficiently described. The primary antibody is referred to only as “rabbit polyclonal anti-human AQP1,” which is not sufficient for methodological reproducibility. The authors should specify the antibody details (for example: manufacturer, catalog number, working dilution, and incubation conditions). In addition, the manufacturer and catalog numbers of the HRP detection system and DAB reagents should also be provided. It would also be important to indicate whether the antibody has been validated for camel tissue or whether cross-reactivity was assumed based on homology.

The antigen retrieval protocol is not clearly described. The manuscript mentions both the use of Cell Marque Trilogy and citrate buffer (pH 6.0), but it is unclear whether these represent two alternative protocols or were used sequentially. The exact antigen retrieval procedure should therefore be clarified, including buffer composition, temperature, and duration.

Negative and positive controls for immunohistochemistry are not described in the manuscript. The authors should specify how the controls were performed (for example: negative control by omission of the primary antibody and the type of tissue used as a positive control to confirm antibody specificity). Such controls are essential for correct interpretation of immunohistochemical staining.

The manuscript presents statistical analysis of “AQP1 expression levels,” yet the method used to obtain quantitative values is not described. The authors should clarify how AQP1 expression was quantified (for example: using an H-score system, optical density measurements, percentage of positive area, or a semi-quantitative scoring approach). Additional methodological details are also necessary, such as the number of microscopic fields analyzed per section, the number of sections analyzed per animal, and whether the analysis was performed in a blinded manner. Without a clearly defined quantification method the statistical analysis cannot be properly evaluated.

The manuscript also reports measurements of corneal layer thickness (epithelium, stroma, and Descemet’s membrane), but the methodology used for these measurements is not described. Based on the current description it is difficult to understand exactly how these measurements were obtained. It should therefore be clarified how the measurements were performed, which software was used for image analysis, how many measurements were obtained per section, and whether the measurements were performed manually or using automated segmentation. Without this information the morphometric analysis cannot be reproduced.

The manuscript repeatedly refers to the use of “assistive AI tools,” but the methodology does not describe any artificial intelligence workflow, algorithm, or software. It would therefore be necessary to clarify what specific software or AI tool was used, how the images were analyzed, whether segmentation or automated measurements were performed, and how the results were validated. If artificial intelligence was not actually used and only conventional digital image analysis was performed, the terminology should be revised accordingly. At present, the role of AI in this study remains unclear and should be described in sufficient methodological detail to ensure reproducibility.

The Materials and Methods section also contains a paragraph describing the use of several electronic databases (Web of Science, IEEE Xplore, PubMed, Scopus, and Google Scholar) and a search strategy for identifying relevant literature. Such a description is typically associated with review articles or systematic literature analyses and does not appear to be directly related to the experimental design of the present study. It would therefore be helpful to clarify the purpose of this section or consider removing it if it is not directly connected with the methodology used in this experimental work.

Figure 8 appears to represent a schematic illustration rather than an experimental micrograph. The legend refers to “green,” “black,” and “red channels,” terminology typically used in fluorescence microscopy. It would therefore be useful to clarify that these colors represent schematic localization of AQP1 rather than imaging channels, and to revise the figure legend accordingly.

Some descriptions of corneal histology appear inconsistent with standard anatomical knowledge. In the Results section the manuscript states that “the amounts of blood vessels were very clear in peripheral corneal parts.” However, the cornea is normally an avascular tissue under physiological conditions. The presence of blood vessels suggests that limbal tissue may have been included in the analyzed samples, so the anatomical boundary between the cornea and the limbus should be clarified and it should be verified whether the peripheral samples contained limbal regions rather than true corneal tissue.

In addition, the Discussion section states that “the corneal endothelium was represented by two to three layers of flattened cells.” This description is inconsistent with the established histological structure of the corneal endothelium, which in mammals is typically composed of a single layer of hexagonal cells. This statement should therefore be verified and revised.

Several issues are also present in the preparation of histological figures. In a number of panels the microphotographs appear to have been resized without maintaining proportional scaling of height and width, which results in images that appear vertically compressed or horizontally stretched. Such distortion may alter the perceived morphology of the tissue and should be corrected by maintaining proportional scaling during figure preparation. In addition, the arrows used to indicate structures are inconsistent in thickness and style, and in some panels arrows of different thickness appear to indicate the same structure, which may confuse the reader. The annotation style should therefore be standardized throughout all figures. Finally, in some panels arrows appear to indicate structures that are likely artifacts rather than anatomical features; for example, in Figure 2C.3 two arrows appear to indicate an artifact rather than a histological structure. All figure annotations should therefore be carefully reviewed to ensure that only relevant anatomical structures are labeled. Overall, improvement of figure contrast, clarity, and annotation consistency would significantly improve the readability of the histological results.

Finally, the term “diameter” should be replaced with “thickness” when referring to corneal layers.

Author Response

Comment No.1. The manuscript states that corneas were obtained from “healthy animals,” but no criteria are provided describing how the health status was determined. It would be important to clarify whether animals underwent veterinary or ophthalmological examination prior to slaughter and to provide basic information about the animals (for example: age, sex, and general condition). In addition, the post-mortem interval and the time between tissue collection and fixation should be specified. Although the manuscript states that the study was approved by the institutional ethics committee of Sadat University, the ethical approval number or protocol identification is not provided. The official approval number and the date of approval issued by the ethics committee should therefore be included.

Answer. Corneas were obtained from clinically healthy animals at the time of slaughter. The animals underwent routine veterinary inspection as part of the standard abattoir health assessment to ensure the absence of systemic disease or ocular abnormalities. Animals showing any signs of ocular pathology, trauma, or systemic illness were excluded from sampling. We have also clarified the post-mortem handling of the tissues. The post-mortem interval between slaughter and corneal excision was immediately after slaughtering, and tissues were immediately placed in fixative after collection to preserve morphological integrity. In addition, the manuscript has been updated to include full ethical approval and basic information details.

Stain                                Species

(Tissue)

Age (Average)

Sex                                              General Conditions

 

H&E                            6 Corneas                  5 years                           3 Males

3 Females

IHC                            6 Corneas                  7 years                           3 Males

3 Females

Under healthy conditions Under healthy conditions

Comment No.2. The cornea was divided into nine regions (central, middle, peripheral; dorsal, ventral, nasal, temporal). However, the manuscript does not explain how these regions were anatomically defined. It would be useful to clarify several methodological aspects: how the boundaries between regions were determined, whether a standardized template or measurement grid was used, and how the orientation of the cornea was preserved during sampling. Without these details it is difficult to evaluate the reproducibility of the regional comparisons.

Answer. The corneal regions were delineated. After excision, the cornea was oriented using the anatomical landmarks of the globe, with the dorsal, ventral and nasal, temporal axes identified prior to dissection. Orientation was preserved during handling by maintaining the original position of the globe and marking the reference axes before tissue processing. The cornea was then subdivided into three concentric zones Central, Middle and Peripheral based on the distance from the corneal apex to the limbus. The central region corresponded to the area surrounding the corneal apex, the middle region represented the intermediate annular zone, and the peripheral region corresponded to the area adjacent to the limbus. Each of these zones was further divided according to the dorsal, ventral, nasal, and temporal orientations to generate the nine predefined sampling regions. The boundaries between regions were determined using proportional measurements relative to the total corneal diameter, and the same orientation procedure was applied to all specimens. These methodological details have now been added to the Materials and Methods section to improve clarity and reproducibility of the regional comparisons.

Comment No.3. Important technical details are also missing from the description of histological processing and morphometric analysis. The manuscript should provide additional information about the histological methodology (for example: the type and manufacturer of the microtome used for sectioning, the criteria used to exclude oblique sections, the number of sections analyzed per animal, and the number of microscopic fields evaluated in each region). Without these details the morphometric measurements cannot be reproduced.

Answer. We agree that providing detailed methodological information is essential to ensure the reproducibility and transparency of morphometric analysis. Accordingly, the Materials and Methods section of the manuscript has been revised to include additional details regarding histological processing and morphometric evaluation. Corneal tissues were fixed in 10% neutral buffered formalin, dehydrated through a graded ethanol series, cleared in xylene, and embedded in paraffin. Serial sections of 4–5 μm thickness were obtained using a rotary microtome. All measurements were performed under identical magnification conditions, and the same analytical procedure was applied to all specimens to maintain consistency. These additional methodological details have now been incorporated into the revised Materials and Methods section to facilitate reproducibility of the histological and morphometric analyses.

Comment No.4 and 5. The immunohistochemical method is insufficiently described. The primary antibody is referred to only as “rabbit polyclonal anti-human AQP1,” which is not sufficient for methodological reproducibility. The authors should specify the antibody details (for example: manufacturer, catalog number, working dilution, and incubation conditions). In addition, the manufacturer and catalog numbers of the HRP detection system and DAB reagents should also be provided. It would also be important to indicate whether the antibody has been validated for camel tissue or whether cross-reactivity was assumed based on homology. The antigen retrieval protocol is not clearly described. The manuscript mentions both the use of Cell Marque Trilogy and citrate buffer (pH 6.0), but it is unclear whether these represent two alternative protocols or were used sequentially. The exact antigen retrieval procedure should therefore be clarified, including buffer composition, temperature, and duration.

Answer. In the revised manuscript, the primary antibody is now fully described. Briefly, immunohistochemical detection of Aquaporin-1 (AQP1) was performed using Rabbit polyclonal anti- human AQP1 antibody (1:1000; catalog no. GB11310, Service bio). Antigen retrieval enhanced by heating in citrate buffer PH (10 mM, pH 6.0). Tissue sections were immersed in the buffer and heated in a water bath at 95–98 °C for 15–20 minutes, followed by gradual cooling to room temperature.

 

The sections were then rinsed into phosphate-buffered saline (PBS) prior to further immunohistochemical processing. Then, 3% hydrogen peroxide was used to block endogenous peroxidase activity following standard immunohistochemical procedures. Horseradish Peroxidase Polymer (HRP)-conjugated secondary detection system (EnVi-sion™+ System-HRP, Dako, Denmark; catalog no. K4003) according to the manufacturer’s instructions for 15 minutes. Then, Substrate/Chromogen Prepare Ready-To-Use DAB substrate solution. Add DAB Chromogen (DAB Substrate Kit, Vector Laboratories, USA; catalog no. SK-4100) Solution (Reagent B1) to DAB Buffer Solution (Reagent B2) and mix the two solutions in a 1:1 ratio with the volume determined by the number of slides to stain. In general, 200 µL of mixed substrate solution is sufficient to cover one tissue slide.

Comment No.6. Negative and positive controls for immunohistochemistry are not described in the manuscript. The authors should specify how the controls were performed (for example: negative control by omission of the primary antibody and the type of tissue used as a positive control to confirm antibody specificity). Such controls are essential for correct interpretation of immunohistochemical staining.

Answer. We now specify that both negative and positive controls were included in each immunohistochemical staining run to verify the specificity and reliability of the antibody reaction. For the negative control, tissue sections were processed following the same immunohistochemical protocol but with omission of the primary antibody. In these sections, the primary antibody was replaced with phosphate-buffered saline (PBS) or normal rabbit serum. This procedure ensured that any observed staining was not due to non-specific binding of the secondary antibody or the detection system. For the positive control, tissue known to express Aquaporin-1 (AQP1) was used to confirm the reactivity of the primary antibody. Sections from kidney tissue, particularly the proximal tubules and vascular endothelium were processed in parallel with the experimental samples using the same staining protocol. The presence of a clear and specific immunoreaction in the positive control confirmed the functionality of the antibody and the effectiveness of the detection system.

Comment No.7. The manuscript presents statistical analysis of “AQP1 expression levels,” yet the method used to obtain quantitative values is not described. The authors should clarify how AQP1 expression was quantified (for example: using an H-score system, optical density measurements, percentage of positive area, or a semi-quantitative scoring approach). Additional methodological details are also necessary, such as the number of microscopic fields analyzed per section, the number of sections analyzed per animal, and whether the analysis was performed in a blinded manner. Without a clearly defined quantification method the statistical analysis cannot be properly evaluated.

Answer. Quantification of AQP1 immunostaining was performed using digital image analysis software (ImageJ or equivalent image-analysis platform). The analysis involved measuring the percentage of positively stained areas and the mean optical density of the DAB signal within the selected fields after background correction. These measurements were used to generate quantitative values representing the relative level of AQP1 immunoreactivity in each region. all measurements were performed under identical imaging parameters, and the analysis was conducted in a blinded manner with respect to the anatomical region and sample identity. The average values obtained from the analysed fields and sections for each animal were then used for statistical comparison among the different corneal regions. These methodological details have now been incorporated into the revised Materials and Methods section to ensure clarity, transparency, and reproducibility of the AQP1 quantification and the subsequent statistical analysis presented in the study.

 

Comment No.8. The manuscript also reports measurements of corneal layer thickness (epithelium, stroma, and Descemet’s membrane), but the methodology used for these measurements is not described. Based on the current description it is difficult to understand exactly how these measurements were obtained.

It should therefore be clarified how the measurements were performed, which software was used for image analysis, how many measurements were obtained per section, and whether the measurements were performed manually or using automated segmentation. Without this information the morphometric analysis cannot be reproduced.

Answer. We agree with the reviewer that the reproducibility of morphometric analysis is essential. We have updated the Materials and Methods section to include a comprehensive description of the corneal layer thickness measurements. Briefly, measurements were performed on digital micrographs using Software. ImageJ via manual linear measurement tools. To ensure accuracy, five random points were measured per section across three non-consecutive sections per animal. Specific details regarding the identification of the epithelium, stroma, and Descemet’s membrane boundaries have been added to the text.

Comment No.9. The manuscript repeatedly refers to the use of “assistive AI tools,” but the methodology does not describe any artificial intelligence workflow, algorithm, or software. It would therefore be necessary to clarify what specific software or AI tool was used, how the images were analyzed, whether segmentation or automated measurements were performed, and how the results were validated. If artificial intelligence was not actually used and only conventional digital image analysis was performed, the terminology should be revised accordingly. At present, the role of AI in this study remains unclear and should be described in sufficient methodological detail to ensure reproducibility.

Answer. Using standard digital image analysis techniques, histological and immunohistochemical images were examined. A light microscope with a digital camera was used to take micrographs under uniform lighting and magnification conditions. ImageJ software (National Institutes of Health, USA) was then used to process and analyze the images. The software's calibrated measuring and thresholding tools were used to quantify immuno-histochemical staining and make morphometric measures. The predetermined corneal regions were manually identified as regions of interest, and all samples were subjected to standardized techniques to collect measurements. This study did not use any machine learning models, automated segmentation techniques, or algorithms based on artificial intelligence. To guarantee uniformity and reproducibility, all analyses were carried out using standard image processing software.

Comment No.10. The Materials and Methods section also contains a paragraph describing the use of several electronic databases (Web of Science, IEEE Xplore, PubMed, Scopus, and Google Scholar) and a search strategy for identifying relevant literature. Such a description is typically associated with review articles or systematic literature analyses and does not appear to be directly related to the experimental design of the present study. It would therefore be helpful to clarify the purpose of this section or consider removing it if it is not directly connected with the methodology used in this experimental work.

Answer. We thank the reviewer for this insightful observation. We agree that the description of the literature search strategy is more appropriate for a review article and does not directly describe the experimental procedures of this study. To improve the clarity and focus of the Materials and Methods section, we have removed the paragraph describing the electronic database searches. References used to support our experimental design remain appropriately cited in the Introduction and Discussion sections.

Comment No.11. Figure 8 appears to represent a schematic illustration rather than an experimental micrograph. The legend refers to “green,” “black,” and “red channels,” terminology typically used in fluorescence microscopy. It would therefore be useful to clarify that these colors represent schematic localization of AQP1 rather than imaging channels, and to revise the figure legend accordingly.

Answer. We thank the reviewer for this important clarification. Figure 8 is indeed a schematic illustration intended to summarize the localized expression of AQP1 across the corneal layers, rather than an experimental micrograph. We acknowledge that the term "channels" was used incorrectly we have revised the Figure 8 legend to replace "channels" with "color markers".

Comment No.12. Some descriptions of corneal histology appear inconsistent with standard anatomical knowledge. In the Results section the manuscript states that “the amounts of blood vessels were very clear in peripheral corneal parts.” However, the cornea is normally an avascular tissue under physiological conditions. The presence of blood vessels suggests that limbal tissue may have been included in the analyzed samples, so the anatomical boundary between the cornea and the limbus should be clarified and it should be verified whether the peripheral samples contained limbal regions rather than true corneal tissue.

Answer. We appreciate the reviewer’s correction regarding corneal vascularization. We acknowledge that the healthy cornea is an avascular structure. The vessels observed and mentioned in the Results section were located at the corneoscleral limbus at this junction, which was included in the peripheral sections to provide complete anatomical reference. and replaced in text by the amounts of blood vessels were very clear in peripheral corneal parts were occasionally observed near the corneal–limbal transition zone. These vessels likely represented extensions of the limbal vascular network rather than true vascularization of the corneal tissues which was normally avascular under physiological conditions. And we suggested receiving nutrients via diffusion from the aqueous humor and the limbal vascular network.

Comment No.13. In addition, the Discussion section states that “the corneal endothelium was represented by two to three layers of flattened cells.” This description is inconsistent with the established histological structure of the corneal endothelium, which in mammals is typically composed of a single layer of hexagonal cells. This statement should therefore be verified and revised.

Answer. We sincerely thank the reviewer for identifying this histological inaccuracy. We acknowledge that the corneal endothelium is a single layer of squamous/hexagonal cells. We have verified our histological sections and confirmed that the endothelium in our samples consists of a single layer of flattened cells.

Comment No.14. Several issues are also present in the preparation of histological figures. In a number of panels the microphotographs appear to have been resized without maintaining proportional scaling of height and width, which results in images that appear vertically compressed or horizontally stretched. Such distortion may alter the perceived morphology of the tissue and should be corrected by maintaining proportional scaling during figure preparation. In addition, the arrows used to indicate structures are inconsistent in thickness and style, and in some panels arrows of different thickness appear to indicate the same structure, which may confuse the reader. The annotation style should therefore be standardized throughout all figures. Finally, in some panels arrows appear to indicate structures that are likely artifacts rather than anatomical features; for example, in Figure 2C.3 two arrows appear to indicate an artifact rather than a histological structure. All figure annotations should therefore be carefully reviewed to ensure that only relevant anatomical structures are labeled. Overall, improvement of figure contrast, clarity, and annotation consistency would significantly improve the readability of the histological results.

Answer. We sincerely apologize for the technical inconsistencies in the figure preparation. We have performed a comprehensive revision of all histological figures to meet the requested standards. Also, the term “diameter” replaced with “thickness” when referring to corneal layers in the text.

 

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript investigates the spatial distribution of aquaporin-1 (AQP1) in the cornea of the dromedary camel using histology, immunohistochemistry, and computer-assisted image analysis. The topic is interesting because the camel eye represents an adaptation to harsh desert environments, and understanding mechanisms of corneal hydration may be relevant for veterinary ophthalmology and comparative biology.

The manuscript presents morphological observations of corneal layers and reports AQP1 immunoexpression in different regions of the camel cornea. However, several aspects of the manuscript require major clarification and improvement before the work can be considered for publication:

First, the Materials and Methods section is difficult to follow and lacks important methodological details. Many steps of the experimental procedure are described in a very unclear way. For example, the immunohistochemistry protocol is written as a sequence of technical instructions rather than a clear scientific description. It is also not clear how many sections were analyzed per sample, how the staining intensity was evaluated, or how the quantitative data were obtained.

Second, the statistical analysis is poorly described. The manuscript mentions the use of one-way ANOVA, but it does not clearly explain what data were analyzed. It is also unclear how AQP1 expression was quantified from the images and how the measurements were performed. Without this information, it is difficult to assess the reliability of the results.

Third, the role of artificial intelligence in the study is not clearly explained. The manuscript repeatedly refers to “assistive AI tools,” but there is no detailed description of the software or the analysis procedure. At the moment, it appears that the AI was mainly used for image measurements, which is common in many standard image analysis programs. The authors should clearly explain what method was used and what makes it an AI-based approach.

Fourth, part of the Materials and Methods section contains information that is not relevant to the study. For example, the manuscript includes a description of database searches and explains what Google Scholar is and why it was used. Such explanations are unnecessary in a research article and should be removed. The Materials and Methods section should focus only on the experimental procedures used in the study.

Fifth, the novelty of the study should be better explained. The manuscript states that this is the first study describing AQP1 expression in the camel cornea, but the introduction does not sufficiently discuss previous research on aquaporins in corneal tissues of other species. A clearer comparison with existing literature is needed.

Finally, the discussion sometimes includes interpretations that go beyond the presented data. The authors frequently suggest that the observed differences are adaptations to desert environments. While this hypothesis is interesting, it is not directly tested in this study and should therefore be discussed more cautiously.

The manuscript contains many grammatical and stylistic errors that make the text difficult to read. The English language should be carefully revised. Some sections of the manuscript are repetitive, especially the descriptions of corneal layers and general corneal physiology. The figure legends are very long and include information that should instead appear in the Materials and Methods section.

 

Author Response

Comment No.1. The Materials and Methods section is difficult to follow and lacks important methodological details. Many steps of the experimental procedure are described in a very unclear way. For example, the immunohistochemistry protocol is written as a sequence of technical instructions rather than a clear scientific description. It is also not clear how many sections were analyzed per sample, how the staining intensity was evaluated, or how the quantitative data were obtained.

Answer. We agree that providing detailed methodological information is essential to ensure the reproducibility and transparency of morphometric analysis. Accordingly, the Materials and Methods section of the manuscript has been revised to include additional details regarding histological processing and morphometric evaluation.

Comment No.2. The statistical analysis is poorly described. The manuscript mentions the use of one-way ANOVA, but it does not clearly explain what data were analyzed. It is also unclear how AQP1 expression was quantified from the images and how the measurements were performed. Without this information, it is difficult to assess the reliability of the results.

Answer. We appreciate the reviewer’s request for clarification on the statistical and quantitative methods. Quantification of AQP1 immunostaining was performed using digital image analysis software (ImageJ or equivalent image-analysis platform). The analysis involved measuring the percentage of positively stained areas and the mean optical density of the DAB signal within the selected fields after background correction. These measurements were used to generate quantitative values representing the relative level of AQP1 immunoreactivity in each region. So, we used one-way AVOVA for measurements of all analyzed data.

Comment No.3. The role of artificial intelligence in the study is not clearly explained. The manuscript repeatedly refers to “assistive AI tools,” but there is no detailed description of the software or the analysis procedure. At the moment, it appears that the AI was mainly used forimage measurements, which is common in many standard image analysis programs. The authors should clearly explain what method was used and what makes it an AI-based approach.

Answer. Using standard digital image analysis techniques, histological and immunohistochemical images were examined. A light microscope with a digital camera was used to take micrographs under uniform lighting and magnification conditions. ImageJ software (National Institutes of Health, USA) was then used to process and analyze the images. The software's calibrated measuring and thresholding tools were used to quantify immuno-histochemical staining and make morphometric measures.

Comment No.4. Part of the Materials and Methods section contains information that is not relevant to the study. For example, the manuscript includes a description of database searches and explains what Google Scholar is and why it was used. Such explanations are unnecessary in a research article and should be removed. The Materials and Methods section should focus only on the experimental procedures used in the study.

Answer. We thank the reviewer for this insightful observation. We agree that the description of the literature search strategy is more appropriate for a review article and does not directly describe the experimental procedures of this study. To improve the clarity and focus of the Materials and Methods section, we have removed the paragraph describing the electronic database searches. References used to support our experimental design remain appropriately cited in the Introduction and Discussion sections.

Comment No.5. The novelty of the study should be better explained. The manuscript states that this is the first study describing AQP1 expression in the camel cornea, but the introduction does not sufficiently discuss previous research on aquaporins in corneal tissues of other species. A clearer comparison with existing literature is needed.

Answer. We appreciate the reviewer’s suggestion to better set the scene our findings within the broader field of Aquaporin research. We have significantly expanded the Introduction and Discussion to provide a comprehensive comparison. We have added a summary of AQP1 expression in humans and other domestic mammals, noting that while the protein's presence in the corneal endothelium is well- documented in these species, its specific spatial density in the camel has never been explored.

Comment No.6. The discussion sometimes includes interpretations that go beyond the presented data. The authors frequently suggest that the observed differences are adaptations to desert environments. While this hypothesis is interesting, it is not directly tested in this study and should therefore be discussed more cautiously.

Answer. We appreciate the reviewer’s insightful comment regarding the interpretation limits of our data. We agree that while the camel’s unique environment provides a unique adaptation for our findings, our study was not designed to test adaptation directly.

 

 

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript investigates the spatial distribution of Aquaporin-1 (AQP1) in the camel cornea using histological and immunohistochemical methods combined with assistive AI tools. The topic is interesting and potentially relevant to veterinary ophthalmology and comparative anatomy.

The manuscript is generally understandable; however, several methodological and conceptual issues need to be addressed before the work can be considered for publication. In particular, the description of the AI methodology, statistical analysis, and immunohistochemical procedures is insufficient.

Major Comments

1. Insufficient description of the AI methodology

The title emphasizes the use of “Assistive AI Applications”, yet the manuscript does not clearly explain:

which algorithm or model was used

which software or platform was employed

what specific task the AI performed (e.g., image segmentation, cell detection, quantification)

how the AI analysis was validated.

Maybe reconsider not including AI in the title.

2. Limited sample size and lack of biological characterization

The study includes 12 camel corneas, but the manuscript does not provide sufficient information about the animals.

The authors should specify:

sex of the animals

approximate age

health status

origin of the samples.

Providing this information is important for assessing biological variability and the reproducibility of the findings.

3. Lack of clear quantitative and statistical analysis

The results mainly describe regional differences in AQP1 distribution; however, it is unclear whether quantitative analysis was performed.

The manuscript should clarify:

whether immunostaining intensity or positive cell counts were quantified

how measurements were obtained

which statistical tests were applied

significance thresholds.

4. Incomplete description of the immunohistochemistry protocol

Important methodological details are missing from the immunohistochemistry section.

The manuscript should include:

antibody name and target species

manufacturer and catalog number

antibody dilution

incubation conditions

detection system used

positive and negative controls.

Providing these details is essential for reproducibility.

5. Overinterpretation in the discussion

Some parts of the discussion propose adaptive mechanisms related to desert environments, but the study does not directly test these physiological mechanisms.

The authors should:

reduce speculative statements

focus on interpretations supported directly by the histological findings

clearly distinguish between results and hypotheses.

8. Please define the aim of the study

Minor Comments

1. English language and grammar

Several sentences require grammatical correction and stylistic improvement.

Author Response

Comment No.1. Insufficient description of the AI methodology. The title emphasizes the use of “Assistive AI Applications”, yet the manuscript does not clearly explain: Which algorithm or model was used, which software or platform was employed, what specific task the AI performed (e.g., image segmentation, cell detection, quantification). how the AI analysis was validated. Maybe reconsider not including AI in the title.

Answer. Using standard digital image analysis techniques, histological and immunohistochemical images were examined. A light microscope with a digital camera was used to take micrographs under uniform lighting and magnification conditions. ImageJ software (National Institutes of Health, USA) was then used to process and analyze the images. The software's calibrated measuring and thresholding tools were used to quantify immuno-histochemical staining and make morphometric measures.

Comment No.2. Limited sample size and lack of biological characterization. The study includes 12 camel corneas, but the manuscript does not provide sufficient information about the animals. The authors should specify: sex of the animals, approximate age, health status, origin of the samples. Providing this information is important for assessing biological variability and the reproducibility of the findings.

Answer. Corneas were obtained from clinically healthy animals at the time of slaughter. The animals underwent routine veterinary inspection as part of the standard abattoir health assessment to ensure the absence of systemic disease or ocular abnormalities. Animals showing any signs of ocular pathology, trauma, or systemic illness were excluded from sampling. We have also clarified the post-mortem handling of the tissues. The post-mortem interval between slaughter and corneal excision was immediately after slaughtering, and tissues were immediately placed in fixative after collection to preserve morphological integrity. In addition, the manuscript has been updated to include full ethical approval (Ethical approval No. VUSC-002-1-26) and basic information details in this table.

Comment No.3. Lack of clear quantitative and statistical analysis. The results mainly describe regional differences in AQP1 distribution; however, it is unclear whether quantitative analysis was performed. And Incomplete description of the immunohistochemistry protocol Important methodological details is missing from the immunohistochemistry section.

Answer. Quantification of AQP1 immunostaining was performed using digital image analysis software (ImageJ or equivalent image-analysis platform). The analysis involved measuring the percentage of positively stained areas and the mean optical density of the DAB signal within the selected fields after background correction. These measurements were used to generate quantitative values representing the relative level of AQP1 immunoreactivity in each region. all measurements were performed under identical imaging parameters, and the analysis was conducted in a blinded manner with respect to the anatomical region and sample identity. The average values obtained from the analysed fields and sections for each animal were then used for statistical comparison among the different corneal regions. These methodological details have now been incorporated into the revised Materials and Methods section to ensure clarity, transparency, and reproducibility of the AQP1 quantification and the subsequent statistical analysis presented in the study. We now specify that both negative and positive controls were included in each immunohistochemical staining run to verify the specificity and reliability of the antibody reaction. For the negative control, tissue sections were processed following the same immunohistochemical protocol but with omission of the primary antibody. In these sections, the primary antibody was replaced with phosphate-buffered saline (PBS) or normal rabbit serum. This procedure ensured that any observed staining was not due to non-specific binding of the secondary antibody or the detection system. For the positive control, tissue known to express Aquaporin-1 (AQP1) was used to confirm the reactivity of the primary antibody. Sections from kidney tissue, particularly the proximal tubules and vascular endothelium were processed in parallel with the experimental samples using the same staining protocol. The presence of a clear and specific immunoreaction in the positive control confirmed the functionality of the antibody and the effectiveness of the detection system.

Comment No.4. Overinterpretation in the discussion. Some parts of the discussion propose adaptive mechanisms related to desert environments, but the study does not directly test these physiological mechanisms.

Answer. We appreciate the reviewer’s insightful comment regarding the interpretation limits of our data. We agree that while the camel’s unique environment provides a unique adaptation for our findings, our study was not designed to test adaptation directly.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The revised manuscript shows clear improvement compared to the initial submission, particularly in the expansion of methodological details and clarification of several previously raised concerns. However, a number of important issues remain unresolved and still limit the scientific clarity and overall quality of the study.

The most significant concern relates to the use of “AI tools.” Despite repeated references throughout the manuscript, it is still not clearly explained how artificial intelligence was actually applied. Based on the current description, the analysis appears to rely on standard ImageJ-based measurements and thresholding, which cannot be considered artificial intelligence in the methodological sense. The authors should either clearly define and justify the use of AI (including software, algorithms, and validation) or revise the terminology accordingly.

In addition, the presentation of histological figures remains below the expected standard. Several issues are noted, including inconsistent and non-standardized scale bars (in some cases multiple scale bars are present within the same image), unnecessary graphical elements such as measurement lines and numerical labels that overload and visually obscure the images, and annotations that are too small to be clearly visible. These aspects significantly reduce the readability and interpretability of the figures. A thorough revision of all figures is strongly recommended, with emphasis on clarity, consistency, and standardization.

Overall, although the manuscript has improved, important methodological and presentation-related ambiguities remain, and further revision is required before it can be considered for publication.

Author Response

Comment No.1. The most significant concern relates to the use of “AI tools.” Despite
repeated references throughout the manuscript, it is still not clearly explained how
artificial intelligence was actually applied. Based on the current description, the analysis
appears to rely on standard ImageJ-based measurements and thresholding, which cannot
be considered artificial intelligence in a methodological sense. The authors should either
clearly define and justify the use of AI (including software, algorithms, and validation) or
revise the terminology accordingly.
Answer. We acknowledge that the term “AI tools” was not sufficiently defined in the original
manuscript and may have led to confusion. Our analysis was conducted using ImageJ in
combination with the Trainable Weka Segmentation plugin, which employs supervised
machine learning algorithms (specifically random forest classifiers) for pixel-based image
classification. This approach goes beyond conventional thresholding by allowing the model to
be trained in user-annotated regions to distinguish between different tissue components and
staining intensities.
In our workflow, representative areas were manually annotated to train the classifier, after which
the trained model was applied to segment AQP1 immunostaining across different samples. The
resulting segmented images were then quantitatively analyzed using ImageJ measurement tools.
To ensure reliability, classifier performance was visually inspected and iteratively refined
through repeated training until consistent segmentation was achieved across samples. To validate
the segmentation results, outputs were visually compared with the original histological images to
confirm accurate detection of AQP1 immunoreactivity. We proved these details in manuscript
with green color.
Comment No.2. In addition, the presentation of histological figures remains below the
expected standard. Several issues are noted, including inconsistent and non-standardized
scale bars (in some cases multiple scale bars are present within the same image),
unnecessary graphical elements such as measurement lines and numerical labels that
overload and visually obscure the images, and annotations that are too small to be clearly
visible. These aspects significantly reduce the readability and interpretability of the figures.
A thorough revision of all figures is strongly recommended, with emphasis on clarity,
consistency, and standardization.
Answer. We sincerely appreciate the reviewer’s detailed and constructive feedback regarding the
quality of our histological figures. We agree that the clarity and standardization of visual data are
paramount for accurate interpretation. In response, we have performed a comprehensive overhaul
of all figures to meet high-standard publication requirements.
All unnecessary graphical overlays, including raw measurement lines and non-essential
numerical labels, have been removed to ensure the underlying morphology is unobstructed and
remains the primary focus. We have replaced small, obscured annotations with high-contrast,
appropriately sized labels and arrows. We have also ensured that the font size and style are
consistent across all panels

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript investigates AQP1 distribution in the camel cornea using histology, immunohistochemistry, and image analysis. The topic is interesting, but the manuscript has a serious conceptual problem. The title claims the use of “assistive AI applications”, however no real artificial intelligence method is presented. The analysis is based on standard ImageJ measurements, thresholding, and manual selection of regions, which cannot be considered AI. This makes the title misleading and does not reflect the actual methodology.

In addition, the Materials and Methods section is still not clear and not fully reproducible. It is not well explained how measurements were performed, how many data points were collected, and what was used as a statistical unit. The statistical analysis is also insufficiently described, with no information about post hoc tests or data independence. The results are mostly descriptive, and the conclusions, especially those related to environmental adaptation, are too strong for the presented data.

For these reasons, I cannot recommend publication of this manuscript in its current form.

Author Response

Comment No.1. This manuscript investigates AQP1 distribution in the camel cornea using
histology, immunohistochemistry, and image analysis. The topic is interesting, but the
manuscript has a serious conceptual problem. The title claims the use of “assistive AI
applications”, however no real artificial intelligence method is presented. The analysis is
based on standard ImageJ measurements, thresholding, and manual selection of regions,
which cannot be considered AI. This makes the title misleading and does not reflect the actual
methodology.
Answer. We agree that the original version of the manuscript did not sufficiently clarify how
artificial intelligence was incorporated, which may have made the title appear misleading. In our
study, image analysis was not limited to conventional thresholding and manual measurements in
ImageJ. Specifically, we employed the Trainable Weka Segmentation (TWS) plugin within
ImageJ, which is based on supervised machine learning algorithms. This approach allows the user
to train a classifier using manually annotated regions, after which the model automatically
performs pixel-based classification across the entire image based on learned features such as
intensity, texture, and edge characteristics. We acknowledge that this represents a form of assistive
or classical machine learning rather than advanced deep learning or fully automated AI systems.
To address this concern, we have revised the manuscript to clearly describe the analytical
workflow, including the training process, feature extraction, and classification steps involved in
Weka segmentation. TWS is a machine learning-based tool that employs a Random Forest
classifier to segment images. Unlike standard global thresholding, which relies on simple pixel
intensity, this AI-driven approach 'learns' from a set of manually annotated training samples to
distinguish between signal and background based on texture, edges, and pixel neighborhoods. (The
details were added into text after revision in green color).
Comment No.2. The Materials and Methods section is still not clear and not fully
reproducible. It is not well explained how measurements were performed.
Answer. Thank you for this valuable comment. We agree that clarity and reproducibility of the
Materials and Methods section are essential. In response, we have thoroughly revised this section
to provide a more detailed and transparent description of the image analysis workflow.
Specifically, we have clarified how measurements were performed using ImageJ, including the
steps for image calibration, region of interest (ROI) selection, and quantification procedures. We
now explicitly describe how ROIs were defined (manual vs. semi-automated), the criteria used for
their selection, and how consistency was maintained across samples. Furthermore, we have
expanded the description of the Trainable Weka Segmentation (TWS) procedure. The revised text
now includes details on the training process (number and type of annotated regions, classes used
for segmentation), the image features selected, and how the trained classifier was applied to all
images. We also clarified how segmentation outputs were converted into quantitative
measurements (area fraction, intensity, or pixel counts). To improve reproducibility, we have
specified the software version, plugins used, and key parameter settings. In addition, we have
indicated that all analyses were performed under standardized conditions to minimize variability.
(The details were added into text after revision in green color).
Comment No.3. How many data points were collected, and what was used as a statistical
unit. The statistical analysis is also insufficiently described.
Answer. We appreciate the reviewer’s request for clarification. We have clarified the sample size
and statistical units in the revised manuscript. A total of 12 corneas from 12 adult camels were
used (n=12), ensuring that each data point represents a unique biological replicate to avoid
pseudoreplication. The samples were divided into two main cohorts: one for morphological
analysis via H&E (n=6; 3 males, 3 females) and one for IHC/AQP1 expression analysis (n=6; 3
males, 3 females). The statistical unit was the individual animal. For thickness measurements and
quantitative IHC (Area Fraction), multiple sections were measured per animal to generate a single
mean value for that individual to compare the differences between the three corneal layers
(Epithelium, Stroma, and Descemet’s membrane/Endothelium), we employed One-way ANOVA
followed by Tukey’s Post-Hoc test for multiple comparisons. This ensures that the specific
variances between layers are statistically validated.". These means were then analyzed using One-
way ANOVA via SPSS to compare the three corneal layers. We have updated the 'Statistical
Analysis' and 'Materials and Methods' sections to reflect these details. (The details were added into
text after revision in green color).

Author Response File: Author Response.pdf

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

The manuscript addresses an interesting topic; however, despite improvements in the revised version, several important methodological and presentation issues remain. These concerns significantly affect the clarity, reproducibility, and interpretation of the results and should be addressed before the manuscript can be considered for publication.

The manuscript places considerable emphasis on the use of “AI tools,” including in the title (“Clinical Spatial Distribution of Aquaporin-1 (AQP1) in Camel Cornea Using Assistive AI Applications”). However, based on the provided methodology, the analysis appears to rely on supervised image segmentation using the Trainable Weka Segmentation plugin within ImageJ. While this represents a useful machine learning-assisted approach, it corresponds to conventional image analysis rather than a fully developed artificial intelligence framework. The authors should therefore more clearly define and justify the use of the term “AI,” including details on the analytical workflow and validation. Alternatively, they may consider moderating this terminology throughout the manuscript, including the title, to more accurately reflect the methodological approach (e.g., “machine learning-based image analysis”).

The Materials and Methods section related to histology and image analysis would benefit from clearer structure and organization. At present, several methodological aspects (including ImageJ-based measurements, morphometric analysis, and machine learning-assisted segmentation) are described across multiple subsections with noticeable overlap and repetition, which makes it difficult to follow the actual analytical workflow.

In particular, it would be important to clearly separate the methodology used for morphometric measurements from that used for quantification of immunopositivity. The authors should explicitly state how many sections per animal and how many microscopic fields per section were analyzed for each type of analysis (morphometry vs. immunohistochemistry). In addition, the definition of a microscopic field should be clarified (i.e., magnification used), as this is essential for reproducibility.

Furthermore, key technical details of image acquisition are missing. The authors should specify the exact model of the Olympus microscope used, the manufacturer and model of the camera, as well as the name and version of the software used for image acquisition and processing (if applicable). These details are necessary to ensure methodological transparency and reproducibility.

A more concise and logically structured presentation—clearly separating histological processing, immunohistochemistry, quantitative image analysis, and machine learning steps—would significantly improve clarity and overall quality of the methodology.

The Results section would benefit from significant improvement in terms of statistical presentation and clarity. Although the authors state that statistical analyses (ANOVA and post-hoc tests) were performed, the results are not adequately supported by statistical evidence.

In particular, figures presenting quantitative data (Figures 6 and 7) do not include any indication of statistical significance (e.g., p-values, asterisks, or letter-based group comparisons), which is standard practice in scientific reporting. As a result, statements describing “higher,” “lower,” or “moderate” expression levels are not sufficiently supported and appear largely descriptive rather than statistically validated.

In addition, the Results section relies heavily on qualitative terminology (e.g., “strong,” “moderate,” “weak” staining) without clearly linking these descriptions to the quantitative data obtained through image analysis. It also remains unclear what the y-axis values (“relative expression levels”) represent, as the method of calculation is not explicitly defined. The figures themselves further contribute to this lack of clarity: in Figures 6 and 7, the y-axis is not properly defined and no measurement units are provided, making interpretation of the numerical values difficult. A similar issue is present in Figure 5, where the meaning of the y-axis values is not clearly explained and no units are indicated. In addition, the labels and annotations (both letters and numbers) in this figure are relatively small and not easily readable, which reduces overall clarity.

Furthermore, certain parts of the Results section contain repeated text, which should be corrected to improve clarity and readability.

The role and necessity of some figures should also be reconsidered. Figure 8 appears to represent a schematic illustration rather than an experimental result, and it is unclear whether it is appropriate to include it within the Results section or whether it is necessary at all. Similarly, Figure 9 raises methodological concerns: the use of “+” symbols suggests a semi-quantitative scoring system; however, such an approach is not described in the Materials and Methods. It is therefore unclear how these scores were obtained and whether they are consistent with the described quantitative image analysis.

Overall, the Results section would benefit from a more concise and data-driven presentation, with clear integration of statistical analysis, explicit reporting of significance, well-defined quantitative metrics, and consistent linkage between quantitative measurements and textual interpretation.

Author Response

Comment. No.1: The manuscript places considerable emphasis on the use of “AI tools,”
including in the title (“Clinical Spatial Distribution of Aquaporin-1 (AQP1) in Camel
Cornea Using Assistive AI Applications”). However, based on the provided methodology,
the analysis appears to rely on supervised image segmentation using the Trainable Weka
Segmentation plugin within ImageJ. While this represents a useful machine learning-
assisted approach, it corresponds to conventional image analysis rather than a fully
developed artificial intelligence framework. The authors should therefore more clearly
define and justify the use of the term “AI,” including details on the analytical workflow
and validation. Alternatively, they may consider moderating this terminology throughout
the manuscript, including the title, to more accurately reflect the methodological approach
(e.g., “machine learning-based image analysis”).
Answer: We thank the reviewer for their rigorous evaluation of our methodology. We have
addressed these concerns by expanding the technical details in the manuscript and providing the
following clarifications:
1. Justification of "AI" Terminology: We respectfully clarify that while basic ImageJ functions
(manual thresholding) are conventional, our study specifically utilized the Trainable Weka
Segmentation (TWS) plugin. This tool is a functional Artificial Intelligence framework based on
Supervised Machine Learning. Unlike standard image analysis, which relies on human-defined
global intensity thresholds, TWS employs a Random Forest classifier. This algorithm was
"trained" using expert-annotated ground-truth samples to learn complex morphological patterns,
edges, and textures (utilizing a high-dimensional feature space including Hessian membranes and
Multi-scale Gaussian filters). This AI-driven approach minimizes operator bias and allows for the
detection of AQP1 distribution patterns that are often indistinguishable via manual thresholding.
We have revised Section 2.10 to explicitly detail this machine-learning architecture, thereby
justifying the term "Assistive AI" in title.
2. Sample Size and Statistical Unit: To ensure biological independence and prevent
pseudoreplication, the individual camel (n=12) was defined as the primary statistical unit. The
study was structured into two distinct cohorts: Morphometric Cohort (n=6; 3 males, 3 females):
For H&E-based thickness measurements. Immunohistochemical (IHC) Cohort (n=6; 3 males, 3
females): For AQP1 expression analysis. For each animal, multiple histological sections were
processed to generate a single representative mean value, ensuring that each data point represents
a unique biological replicate.
3. Enhancement of Statistical Description: The "Statistical Analysis" section (Section 2.11) has
been completely rewritten for clarity. We have specified the use of IBM SPSS Statistics (Version
28.0) and confirmed that data were analyzed using One-way ANOVA followed by Tukey’s
Honest Significant Difference (HSD) post-hoc test for pairwise comparisons between the three
corneal layers. All results are now consistently reported as Mean ± Standard Deviation (SD) with
significance set at P < 0.05.
Comment. No.2: The Materials and Methods section related to histology and image analysis
would benefit from clearer structure and organization. At present, several methodological
aspects (including ImageJ-based measurements, morphometric analysis, and machine
learning-assisted segmentation) are described across multiple subsections with noticeable
overlap and repetition, which makes it difficult to follow the actual analytical workflow.
Answer: We agree with the reviewer’s assessment that the methodological workflow required
better structural integration. In the revised manuscript, we have completely reorganized the
Materials and Methods to eliminate redundancy and provide a clear, chronological analytical
pipeline.
Comment. No.3: In particular, it would be important to clearly separate the methodology
used for morphometric measurements from that used for quantification of
immunopositivity. The authors should explicitly state how many sections per animal and
how many microscopic fields per section were analyzed for each type of analysis
(morphometry vs. immunohistochemistry). In addition, the definition of a microscopic field
should be clarified (i.e., magnification used), as this is essential for reproducibility.
Answer: We have revised the description of our histological and morphometric protocols
(Section 2.3) to provide the specific details requested. We have explicitly stated the criteria for
section selection, focusing on perpendicular orientation to ensure measurement accuracy.
Furthermore, we have clarified our sampling depth: we analyzed 5 non-consecutive sections per
animal and 5 microscopic fields per section at 100x magnification. This standardized approach
across our morphometric cohort (n=6) ensures that the data used for the subsequent One-way
ANOVA is robust, reproducible, and representative of the biological variation within the species.
Also, we have revised (Section 2.6) to provide a granular description of the AQP1 quantification
process. We have explicitly separated the AI-assisted segmentation protocol from the general
image processing steps. As requested, we clarified that IHC analysis was conducted at 400x
magnification to ensure precise pixel-level classification by the Random Forest engine. We have
also detailed the sampling density—analyzing 5 non-consecutive sections per animal and 5
microscopic fields per section—totaling 150 fields for the IHC cohort (n=6). The use of a
blinded examiner and an iterative training-to-inference AI workflow further ensures the
objectivity and reproducibility of our spatial distribution data.
Comment. No.4: Furthermore, key technical details of image acquisition are missing. The
authors should specify the exact model of the Olympus microscope used, the manufacturer
and model of the camera, as well as the name and version of the software used for image
acquisition and processing (if applicable). These details are necessary to ensure
methodological transparency and reproducibility.
Answer: We fully agree that technical transparency regarding hardware is essential for
reproducibility. We have updated Section 2.7 to include the precise specifications of our imaging
system. Micrographs were captured using an Olympus [Insert Model, BX53] upright
microscope equipped with an Olympus [Insert Camera Model, DP27] digital camera. Image
acquisition was managed via the Olympus cell Sens [Insert Version, Standard 1.18] software
suite. All images were saved in uncompressed TIFF format to maintain pixel integrity for the
subsequent AI-assisted analysis in ImageJ.
Comment. No.5: A more concise and logically structured presentation—clearly separating
histological processing, immunohistochemistry, quantitative image analysis, and machine
learning steps—would significantly improve clarity and overall quality of the methodology.
Answer: We appreciate this insightful suggestion to enhance the readability of our methodology.
We have completely restructured the Materials and Methods section to follow a linear, logical
progression that separates the physical laboratory procedures from the computational analysis.
Comment. No.6: The Results section would benefit from significant improvement in terms
of statistical presentation and clarity. Although the authors state that statistical analyses
(ANOVA and post-hoc tests) were performed, the results are not adequately supported by
statistical evidence.
Answer: We appreciate this insightful suggestion, and we have completely restructured the result
section to ensure that it is adequately supported by statistical evidence.

Author Response File: Author Response.pdf