Ophthalmic Imaging in Diabetic Retinopathy and Diabetic Macular Edema: Key Findings and Advancements
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe is a nice review on Ophtalmic Imaging Finding in Diabetic Retinopathy and DME. It provides valuable insight into structural an functional relation of multimodal imaging
The manuscript is clear, well-written and pleasant to read. Sections of OCT-B, OCT-A, FAF, FA and UWFI are structured .
- I appreciate reading section on qualitative OCT-B parameters (DRIL, HRF, HCF cystoid spaces) which is insightful using recent data of the literrature. The authors mentionned strengh and limits of each imaging in daily practice (ie, artefacts in OCT-A)
- Tables, figures are well presented
- It is not easy to find imperfections in this manuscript, I suggest minors details
- AI and deep learning are mentionned. Readers would appreciate to learn more details on sensitivity and specificity in detecting DME with AI
- Legends of image on page 6 is lacking (arrow for VMIA)
- Numbering of figures is lacking
Author Response
a) AI and deep learning are mentioned. Readers would appreciate to learn more details on sensitivity and specificity in detecting DME with AI
The sensitivity and specificity in detecting DME with AI has now been added in the conclusion section.
b) Legends of image on page 6 is lacking (arrow for VMIA)
Thanks for your comment. Legend has been added for Fig. 4. The earlier image for Fig. 4 has been replaced with a new image showing a detailed description of the intraretinal cysts. Reference has also been made to this in the paper. Please see page lines 165-166.
c) Numbering of figures is lacking
We appreciate the reviewer bringing this to our attention and the figures are now number and cited in the text. There are 6 figures in total.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe authors have submitted a manuscript aiming to describe key imaging findings and their advancements in DR and DME.
Unfortunately, there are significant issues with this current manuscript due to problems with nearly all the figures, imbalance in the amount of writing dedicated to each imaging modality, lack of description of the latest advancements in many of the imaging modalities and many factual inaccuracies which are not in line with general / current thinking or consensus guidelines.
MAJOR ISSUES with figures:
1. Figure on page 3: Figure not numbered; No reference to figure in main text; No figure legend describing patient details and diagnosis (is this in the context of DME with other DME findings), other findings in figure not described (outer retinal atrophic changes with EZ and RPE disturbances atrophy). No associated en face image to show which eye and where the slice is taken from- where is fovea?
2. Figure on page 4: Figure not numbered; No reference to figure in main text; No figure legend describing patient details and diagnosis (there is an isolated hyper-reflective focus- is this in the context of a patient with DR / DME?). No associated en face image to show which eye and where the slice is taken from.
3. Figure on page 5: Figure not numbered; No reference to figure in main text; No figure legend describing patient details and a diagnosis (is this from an otherwise healthy retina with an isolated IRC?). No associated en face image to show which eye and where the slice is taken from.
4. Figure on page 6: Figure not numbered; No reference to figure in main text; No figure legend describing patient details and a diagnosis. No description of type of finding and why its shown (there is no major vitreomacular traction here; no clear disruption of retinal architecture). Why is this shown? There are two points of vitreomacular adhesions- so??
5. Figure on page 8: TOTALLY INCORRECT description as shown. This is a tractional retinal detachment secondary to advanced PDR due to significant fibrovascular complications until proven otherwise. How is this diabetic macular edema unless an OCT scan is shown with it? The few macular exudates could be sequelae of old DME? No reference to figure in text; No figure number or figure legend.
6. Figure on page 9: No figure number of reference to figure in text. No grading of DR in this image based on descriptions in text. No annotations.
7. Figure on page 11: No figure number or reference to figure in text. No annotations. Of all the useful information provided by OCT-A, why is an apparent image of 'macrophage-like cells' chosen? Where are they in the image? There is no annotation.
8. An major issue throughout the manuscript is reference to grading systems that are not mentioned in the text, with no reference to a consensus working group. e.g. line 171, a grading of disruption to ELM and EZ (which system is this referring to?); line 174- which grading system is this in reference to integrity of EZ- please name it in the text.
9. Paragraph about choroid starting on line 195: This entire section / paragraph adds no valuable information for the reader and is disorganized with no real message. Please remove it.
10. Colour fundus photography: Would it not be useful in a review paper of imaging to insert an annotated figure demonstrating NPDR stages and PDR as described to in the text rather than the random photo showing complications due to advanced PDR (tractional retinal detachment)?
11. FAF: There is too much waffle about this modality without mentioning the basics- that FAF is useful to investigate RPE stress (hyperautofluorescence) vs RPE death (hypoautofluorescence) and its uses in DR / DME. This is an imaging modality that is much less important than FFA for diabetic patients, yet the amount of information described is disproportionate.
9. OCT-A: There is just too little information on OCT-A in this section and on its most recent advances. This section must be enhanced with factually accurate descriptions- "vascular blood flow within all layers of the retina": OCT-A is organized based on SVC/DVC/AC/choriocapillaris/choroid and these are useful to extrapolate to retinal layers and issues. There is no mention of widefield OCT as a latest advance. Out of all the figures chosen, why provide one of something in infancy- "macrophage-like cells". How is this useful for the reader?
10. FFA: For such an important investigation that is used on near-daily basis in specialist retina clinics, there is too much information here on the technique itself an not enough on imaging findings. Widefield FFA is now used routinely to judge peripheral retinal ischaemia, to direct PRP treatment (FFA-directed PRP), for focal treatment.
11. Line 223: Why is cotton wool spots described as 'soft exudates' within the text. CWS are infarction due to disruption in axoplasmic flow? Where is this information being obtained from?
12. Please ensure your manuscript and figures are reviewed by a retina specialist before submitting.
MINOR ISSUES:
Line 79: This should be DRIL rather than DR15
Line 227: This should be PDR rather than PMR
Author Response
Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections in the re-submitted files.
MAJOR ISSUES with figures:
- Figure on page 3: Figure not numbered; No reference to figure in main text; No figure legend describing patient details and diagnosis (is this in the context of DME with other DME findings), other findings in figure not described (outer retinal atrophic changes with EZ and RPE disturbances atrophy). No associated en face image to show which eye and where the slice is taken from- where is fovea?
Response: Figure 2 has been numbered and now is mentioned in the main text of the manuscript. Figure legend has now been provided for Figure 2.
- Figure on page 4: Figure not numbered; No reference to figure in main text; No figure legend describing patient details and diagnosis (there is an isolated hyper-reflective focus- is this in the context of a patient with DR / DME?). No associated en face image to show which eye and where the slice is taken from.
Response: Figure 3 has been numbered and now is mentioned in the main text of the manuscript. Figure legend has now been provided for Figure 3. The hyperreflective foci was in the context of DR. The figure legend now specifies this. Also, please see page 4 lines 96-101. An associated en face image has now been added.
- Figure on page 5: Figure not numbered; No reference to figure in main text; No figure legend describing patient details and a diagnosis (is this from an otherwise healthy retina with an isolated IRC?). No associated en face image to show which eye and where the slice is taken from.
Response: Figure 4 has been numbered and now is mentioned in the main text of the manuscript. The new Figure 4 shows a DME patient with intraretinal cysts. Figure legend has been provided. Please see also page 5 lines 161-162 which describes the corresponding Figure 4.
- Figure on page 6: Figure not numbered; No reference to figure in main text; No figure legend describing patient details and a diagnosis. No description of type of finding and why its shown (there is no major vitreomacular traction here; no clear disruption of retinal architecture). Why is this shown? There are two points of vitreomacular adhesions- so??
Response: Thanks for your comments. Figure 5 has now been removed from the manuscript. (I suggested in the paper that this figure be removed since we do not have a good VMIA image, because all diseased are DME only.
- Figure on page 8: TOTALLY INCORRECT description as shown. This is a tractional retinal detachment secondary to advanced PDR due to significant fibrovascular complications until proven otherwise. How is this diabetic macular edema unless an OCT scan is shown with it? The few macular exudates could be sequelae of old DME? No reference to figure in text; No figure number or figure legend.
Response: Figure 5 has been removed and we apologize as it shows advanced PDR with fibrovascular membranes but DME is not clearly shown.
- Figure on page 9: No figure number of reference to figure in text. No grading of DR in this image based on descriptions in text. No annotations.
Response: Figure 4 has been numbered and now is mentioned in the main text of the manuscript. Figure legend has now been provided for Figure 5 with detailed descriptions.
- Figure on page 11: No figure number or reference to figure in text. No annotations. Of all the useful information provided by OCT-A, why is an apparent image of 'macrophage-like cells' chosen? Where are they in the image? There is no annotation.
Response: We apologize for not making this clearer in the text; however, Macrophage-like cells (MLCs) and hyperreflective foci (HRF) observed on optical coherence tomography (OCT) are emerging as important biomarkers in (DME, reflecting underlying inflammation and immune activation. MLCs are visualized at the vitreoretinal interface (VRI), particularly near the internal limiting membrane (ILM), using en face OCT and swept-source OCT angiography (SS-OCTA). These cells are believed to include activated microglia and hyalocytes, both of which originate from the monocyte-macrophage lineage. MLC density is significantly increased in eyes with DME and proliferative diabetic retinopathy (PDR) compared to non-DME and non-proliferative DR (NPDR) eyes. MLCs appear larger and more morphologically active in DME, suggesting heightened inflammatory activity. MLC density correlates negatively with vessel density in the deep capillary plexus, indicating a link between inflammation and microvascular compromise.
These cells may contribute to the breakdown of the inner blood-retinal barrier (iBRB), promoting vascular leakage and edema formation [PMID: 36726090]. Figure legend showing detailed description of Figure 6 has been provided.
- An major issue throughout the manuscript is reference to grading systems that are not mentioned in the text, with no reference to a consensus working group. e.g. line 171, a grading of disruption to ELM and EZ (which system is this referring to?); line 174- which grading system is this in reference to integrity of EZ- please name it in the text.
This was a system devised by the authors of this paper, with no formal name. This system has shown a significant correlation to logMAR visual acuity if utilized. We have updated these lines to clarify that this is a proposed system.
- Paragraph about choroid starting on line 195: This entire section / paragraph adds no valuable information for the reader and is disorganized with no real message. Please remove it.
Response: This paragraph has been removed as requested by the reviewer.
- Colour fundus photography: Would it not be useful in a review paper of imaging to insert an annotated figure demonstrating NPDR stages and PDR as described to in the text rather than the random photo showing complications due to advanced PDR (tractional retinal detachment)?
Response: Thanks for your comment. The advanced PDR image showing the tractional retinal detachment has been removed. We agree with your comment. Hence, a new figure 1 showing the various ETDRS grading has now been shown in the paper.
- FAF: There is too much waffle about this modality without mentioning the basics- that FAF is useful to investigate RPE stress (hyperautofluorescence) vs RPE death (hypoautofluorescence) and its uses in DR / DME. This is an imaging modality that is much less important than FFA for diabetic patients, yet the amount of information described is disproportionate.
Response: We have condensed the section on FAF and included the important point raised by the reviewer regarding RPE stress (hyperautofluorescence) vs RPE death (hypoautofluorescence).
- OCT-A: There is just too little information on OCT-A in this section and on its most recent advances. This section must be enhanced with factually accurate descriptions- "vascular blood flow within all layers of the retina": OCT-A is organized based on SVC/DVC/AC/choriocapillaris/choroid and these are useful to extrapolate to retinal layers and issues. There is no mention of widefield OCT as a latest advance. Out of all the figures chosen, why provide one of something in infancy- "macrophage-like cells". How is this useful for the reader?
Response: As requested the section on OCT-A has been improved. The importance of imaging is not only to identify pathology to prompt an accurate and timely diagnosis but also to provide insight into the pathogenesis of the disease and its mechanism. To this end the authors sought to include the section on macrophage like cells to emphasize the key role of inflammation in DME pathogenesis.
- FFA: For such an important investigation that is used on near-daily basis in specialist retina clinics, there is too much information here on the technique itself an not enough on imaging findings. Widefield FFA is now used routinely to judge peripheral retinal ischaemia, to direct PRP treatment (FFA-directed PRP), for focal treatment.
Response: We appreciate this point and have reduced the description of the technique and added a section on widefield FFA.
- Line 223: Why is cotton wool spots described as 'soft exudates' within the text. CWS are infarction due to disruption in axoplasmic flow? Where is this information being obtained from?
Response: We apologize for this error and it was an oversight during the editing of the paper. We have now corrected it.
Author Response File: Author Response.pdf
Reviewer 3 Report
Comments and Suggestions for AuthorsAuthors Akanksha Malepati, et.al in the Review entitled ‘Ophthalmic Imaging Findings in Diabetic Retinopathy and Diabetic Macular Edema: Key Findings and Advancements’ have presented a comprehensive guide for DR and DME imaging that can serve as an effective diagnostic tool. All the contemporary retinal imaging techniques have been deeply discussed with the pros and cons in its application. However, to deliver the compact knowledge from this review to the research audience, it will be good to inculcate some minor edits/inclusions as follows:
Lines 422-426: ‘Conclusion’:
- It will be good to mention in concise about the emerging imaging biomarkers that are currently under investigation for DR and DME and that could be improvised with the intervention of AI / deep learning models.
- To mention in concise on how would deep learning models contribute to improved detection and diagnosis of DR and DME in ophthalmic imaging, especially with regard to the UWFI technique.
Lines: 260-263:
- Has there been a significant reported change in the detection rate of neovascularization and pre-retinal hemorrhages with UWFI compared to the traditional methods? (any study indicating this analysis can be referenced or to be summarized in the section).
Author Response
Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted/in track changes in the re-submitted files
a) It will be good to mention in concise about the emerging imaging biomarkers that are currently under investigation for DR and DME and that could be improvised with the intervention of AI / deep learning models.
We have added a few image markers that have been areas of interest with deep learning models to the conclusion section.
b) To mention in concise on how would deep learning models contribute to improved detection and diagnosis of DR and DME in ophthalmic imaging, especially with regard to the UWFI technique.
We have added a concise section of the sensitivity and specificity of AI in detecting DME, regarding CFP, OCT, and UWFI ophthalmoscope images.
c) Lines: 260-263: Has there been a significant reported change in the detection rate of neovascularization and pre-retinal hemorrhages with UWFI compared to the traditional methods? (any study indicating this analysis can be referenced or to be summarized in the section).
We have added the reported change in detection rate of neovascularization in UWFI compared to traditional methods.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
Comments and Suggestions for AuthorsThere have been improvements to the manuscript- thanks to the authors.
I strongly recommend the following changes (not additional- relating to previous comments) to improve the manuscript further so there is no ambiguity and the information is fully supported by evidence to be in a literature review:
1. Figure 1: Part e shows treated PDR with (what looks like) punctate chorioretinal scarring from PRP scatter laser. There should be a fundus photo included of active, untreated, PDR (with NVE and/or NVD). The photos, in general, should be larger to delineate microaneurysms, dots, blots, exudates. At the current size, it is very difficult to see these details.
2. Figure 3: This figure tries to demonstrate hyper-reflective foci associated with lipid exudates. I'm not sure why the authors chose a photo and corresponding OCT slice with a solitary hyper-reflective focus? Exudates are typically present in groups / clusters / circinates and associated with macular edema (when a diagnosis of clinically-significant macular edema is made). Please chose a more suitable photo/scan that is typical of exudates.
3. Figure 6: Please remove this figure. It is fine to speculate in the main text about macrophage-like cells as possible biomarkers, but on the figure itself, the (approximately) large 1mm lesions pointed to with arrows are not macrophage-like cells or aggregates of cells, and there is not enough evidence that I'm aware of that supports this. Hyalocytes / retinal microglia / choroidal macrophages do not appear like this on scans. For example, in an AMD context, the histological work of Florian Sennlaub (Paris) and Christine Curcio (Alabama) show that antigen-presenting cells appear with RPE / pigment migration. This is a literature review, so if this figure were to be included, it would need back up with histology. I would recommend the authors instead include OCT-A slices showing areas of neovascularisation or retinal non-perfusion or enlarged FAZ (what OCT-A is realistically used for). OCT-A is even used for anterior segment imaging (NVI in PDR), but no information of imaging is provided for this.
Author Response
Thank you very much for taking the time to review this manuscript. Please find the detailed responses below and the corresponding revisions/corrections highlighted/in track changes in the re-submitted files.
Figure 1: Part e shows treated PDR with (what looks like) punctate chorioretinal scarring from PRP scatter laser. There should be a fundus photo included of active, untreated, PDR (with NVE and/or NVD). The photos, in general, should be larger to delineate microaneurysms, dots, blots, exudates. At the current size, it is very difficult to see these details.
Authors’ response: Thanks for your comment. We have replaced part e of Figure 1 with a fundus photo of an active untreated PDR. We also replaced parts b and c of Figure 1 with images that show the features of mild and moderate NPDR, respectively. Arrows were placed in part b to delineate areas of dot and blot hemorrhages on the figure. Part c also shows features consistent with mild NPDR. We have also enlarged Figure 1.
Figure 3: This figure tries to demonstrate hyper-reflective foci associated with lipid exudates. I’m not sure why the authors chose a photo and corresponding OCT slice with a solitary hyper-reflective focus? Exudates are typically present in groups / clusters / circinates and associated with macular edema (when a diagnosis of clinically-significant macular edema is made). Please chose a more suitable photo/scan that is typical of exudates.
Authors’ response: Thanks for your comment. We have now included a new image for Figure 3 that shows clusters of hard exudates associated with macular edema. We show both the color fundus and corresponding OCT images.
Figure 6: Please remove this figure. It is fine to speculate in the main text about macrophage-like cells as possible biomarkers, but on the figure itself, the (approximately) large 1mm lesions pointed to with arrows are not macrophage-like cells or aggregates of cells, and there is not enough evidence that I’m aware of that supports this. Hyalocytes / retinal microglia / choroidal macrophages do not appear like this on scans. For example, in an AMD context, the histological work of Florian Sennlaub (Paris) and Christine Curcio (Alabama) show that antigen-presenting cells appear with RPE / pigment migration. This is a literature review, so if this figure were to be included, it would need back up with histology. I would recommend the authors instead include OCT-A slices showing areas of neovascularization or retinal non-perfusion or enlarged FAZ (what OCT-A is realistically used for). OCT-A is even used for anterior segment imaging (NVI in PDR), but no information of imaging is provided for this.
Authors’ response: Thanks for your detailed feedback. As suggested, we have now included a new Figure 6 consisting of OCTA images showing areas of capillary non perfusion at the fovea (enlarged FAZ), perifoveal, and mid-peripheral retinal regions.
Author Response File: Author Response.pdf
Round 3
Reviewer 2 Report
Comments and Suggestions for AuthorsThanks to the authors for addressing the comments and uploading the revised figures. The review now reads (and looks) improved with no ambiguity or controversial elements in the content. There are some remaining changes which are required: 1) I would strongly recommend placing the text on Colour Fundus Photography (section 2.2) BEFORE Spectral-Domain Optical Coherence Tomography (SD-OCT), (section 2.1). This is not only due to figure 1 using colour fundus images, but also that the OCT biomarkers described in the text and OCT figures are for structural correlates in the fundus photos. 2) In Figure 6, a scale bar is required (or overlay with measuring tool of area) to demonstrate the authors' comment in the legend that this is an enlarged FAZ. As the authors are aware, if FAZ is greater than 0.5mm2, this generally constitutes DMI. Therefore, if the FAZ area in Figure 6 is not greater than 0.5mm2, please specifically state this and/or why its greater than 'normal' FAZ. 3) Please make reference to ELM integrity as a prognostic biomarker in DME.
Author Response
Thank you very much for taking the time to review this manuscript. Please find the detailed responses below.
1) I would strongly recommend placing the text on Colour Fundus Photography (section 2.2) BEFORE Spectral-Domain Optical Coherence Tomography (SD-OCT), (section 2.1). This is not only due to figure 1 using colour fundus images, but also that the OCT biomarkers described in the text and OCT figures are for structural correlates in the fundus photos.
This has been reordered as suggested by the reviewer. Thank you very much for this suggestion.
2) In Figure 6, a scale bar is required (or overlay with measuring tool of area) to demonstrate the authors' comment in the legend that this is an enlarged FAZ. As the authors are aware, if FAZ is greater than 0.5mm2, this generally constitutes DMI. Therefore, if the FAZ area in Figure 6 is not greater than 0.5mm2, please specifically state this and/or why its greater than 'normal' FAZ.
Thank you for your comment. We have now placed a scar bar for Figure 6. We also used the lasso tool to demarcate the measured FAZ area. Finally, we also stated the size of the FAZ which is 0.75 mm2.
3) Please make reference to ELM integrity as a prognostic biomarker in DME.
Thank you for your comment. This has been included now in lines 305-306.
