Accuracy of Image-Based Automated Diagnosis in the Identification and Classification of Acute Burn Injuries. A Systematic Review

Round 1

Reviewer 1 Report

The authors present a systematic review on the accuracy of image-based classification burns.

The idea of objective and/or remote assessment of burn wounds is found throughout the literature with many studies addressing this issue. Consequently . compiling reviews have been published. Therefore this present systematic has to be considered confirmative.

To improve the value of the manuscript ( as compared to other “reviews”) the authors should precisely follow the PRISMA checklist when structing their manuscript. While they present the PRISMA flow chart, the abstract does not meet the criteria of PRISMA; the selection process does not describe the exclusion criteria of the 3 excluded reports. 

In addition the authors should precisely describe whether the included papers  followed CONSORT criteria. They may give these data in a Table.

 

The major aspect which is not clearly described in the review is the comparator for measuring positive predictive value and final accuracy.  Was the comparator the actually observed clinical outcome or was this the subjective assessment of pictures by a clinician.

The authors state that online pictures were used. In this case no clinical outcome can be used to compare for accuracy. This aspect should  be discussed as this is a major problem in many studies on burn wound assessment. Every burn even a superficial burn can be grafted and will heal, but the question which burn will not require grafting is not answered, but that is the most  important issue.

Therefore true accuracy can only be evaluated by analysis of the excised tissue or by not grafting all wounds and observe outcome (latter ethically difficult). Another possibility would be LDI as a comparator. Did any cited paper include this?

 

Furthermore, the authors should discuss or mentioned in the introduction that clinical burn wound assessment is no by vision but rather by pain and perfusion. This is in sharp contrast to dermatological disease e.g., Melanoma where appearance (ABCD criteria) is the main discriminator.

Comments for author File: Comments.pdf

Author Response

The authors would like to thank the reviewer for his time and thoughtful comments. We have taken each of them into consideration, and hope our manuscript is stronger now. Please find our responses to each comment in the attached document.

Author Response File: Author Response.pdf

Reviewer 2 Report

I think that the topic is of interest and worth to be published; I have only one or two minor issues: What are the limitations of the study? These are not mentioned and discussed! What are the requested next steps? What kind of research isneeded? Please state and discuss.

Author Response

We thank the reviewer for their time and comments. We have added quite a large amount of information to the methods and review section which should assist in answering the reviewer's comments.

Specifically, an additional section was added in order to clarify even more what were the limitations of the study and what were the future research needed.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have sufficiently addressed all issues.

I have no further recommendations or open questions.

Some references are still missing in the manuscript.