Importance of Electrode Selection and Number in Reconstructing Standard Twelve Lead Electrocardiograms

Round 1

Reviewer 1 Report

The manuscript by Butchy is very interesting, but it presents some issues that have to be solved. Following, some comments:

•                   Format of the paper should be revised: (1) “ie” should be substituted with “i.e.”; (2) units should be reported according to SI; (3) format does not fit the journal guidelines.

•                   Abstract should contain background and quantitative results. I suggest revising the abstract.

•                   Figure one should be clarified. This is not the standard placement of 10 electrodes for the standard 12-lead ECG recording neither for the standard configuration neither for the Mason-Likar configuration. I strongly suggest correcting the figure.

•                   Database should be presented in a separate section. Moreover, which are the anatomical features of these patients? Which is the technical setting of acquisition of the considered ECG?

•                   About the preprocessing, why did the authors select a cut-frequency of 2Hz? It is not used in standard ECG signal processing, and this can delete all low-frequency content of the signal. Moreover, which is the order of the filter? Did the author apply a bidirectional filtering?

•                   Statistics should be described in a separate section.

•                   Did the authors evaluate the dependency of the method to signal to noise ratio and sample frequency of the recording? In my opinion, these features may interfere with the results. I suggest inserting a robustness analysis to these features by using a simulated dataset.

•                   I suggest considering real clinical scenario that may benefit from this technology. One can be the use of these algorithms in case of portable and wearable devices (e.g., DOI 10.1109/TCE.2023.3237715; DOI 10.3390/s20123570; DOI 10.1007/978-981-19-5221-0_5). I suggest at least discussing this scenario in the discussion section.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

This article focuses on the use of less than twelve leads for reconstruction of the classic two-lead ECG. The work is interesting, but I'm not sure about it's novelty to the research group. 

1) Try shortening the title of the article. A good and punchy title has less than eight words without prepositions.

2) The images could be of better quality and preferably vector.

3) I would expect more description of the neural network used.

4) I miss novelties at work. In the introduction of the thesis, clearly state the novelty.

5) From the results of the paper, I don't think the reader is able to reproduce your experiment. I would have expected a better description and demonstration.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The manuscript still presents some issues that need be solved:

- Units are still not in SI format.

- As authors reported in the answer to reviewers, the SNR interferes with the results. I strongly suggest including the robustness analysis in this paper and not only in the answer to Reviewers.

- In my opinion, clinical utility of the proposed method in the real clinical scenario should be discussed in more detail.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors have corrected the article according to my recommendations. I now recommend it for acceptance.

Author Response

Reviewer recommended for acceptance

Round 3

Reviewer 2 Report

The authors have corrected the article according to my recommendations and those of other reviewers. I now recommend it for acceptance.

Author Response

No Comments Provided