Assessment of Compressed and Decompressed ECG Databases for Telecardiology Applying a Convolution Neural Network

Round 1

Reviewer 1 Report

1、There is a numeric error 02.61 in line 23 of the text.

2、The innovation of the paper is not well summarized，suggest  rewrite it.

3、The forms and formulas in the paper are not standardized.

4、In Section 5，there's an extra half parenthesis on line 459.

5、In Section: Compression & Decompression Outcomes, the method of compressed and decompressed ECG databases is not detailed, recommended extension description.

Author Response

RESPONSES OF THE REVIEWER’S COMMENTS

August 19, 2022

Paper ID: electronics-1875224

Title: Assessment of Compressed and Decompressed ECG Databases for Telecardiology Applying Convolution Neural Network

Journal: Electronics, MDPI

Special issue-Computer science and Engineering

Dear Editors and Reviewers

 

Thank you so much for allowing revision for our paper. We have given the response to the reviewer’s comments and suggestions below. Besides, we also revised everything suggested carefully in the paper. We believe that the paper is ready for publication now. Thank you so much for your assistance.

Sincerely,

Authors

Brazil-India

Author Response File: Author Response.pdf

Reviewer 2 Report

The authors of this paper propose a system for compressing ECG signals based on DCT/IDCT. The data from the Physionet database is used to train a CNN algorithm for diagnosing ECG signals. However, the relative between these subsystems is not clear in this paper, and the authors should explanations more details on the following issues:

 

1. The author must explain the design method of the quantization coefficient after the DCT process of the ECG signal in this study.

 

2. Please explain this article's compression codec method. DCT/IDCT and quantization are not data compression methods but only preprocessing.

 

3. Please explain the effect of the quantization coefficient designed in this study on SNR and compression rate.

 

4. Please explain how the ATMega328P microprocessor in this paper runs the system in real-time, including the compression codec algorithm and the CNN algorithm for ECG diagnosis.

 

5. Please explain the influence of ECG signal features on CNN detection. For example, the ECG signal using the Physionet database has a different sampling rate than the author using ATMega328P to sample the ECG signal from AD8232.

 

6. Can the proposed CNN algorithm run on the ATMega328P microprocessor? Please describe the method of implementation and assessment. Suppose the compressed signal is transmitted to another device with high computing power for analysis. In that case, the author must describe the transmission method and the technical details of the apparatus that runs the CNN algorithm.

 

7. Please explain the results for TP, TN, FP, and FN and the unit in which the target was detected.

Author Response

RESPONSES OF THE REVIEWER’S COMMENTS

August 19, 2022

Paper ID: electronics-1875224

Title: Assessment of Compressed and Decompressed ECG Databases for Telecardiology Applying Convolution Neural Network

Journal: Electronics, MDPI

Special issue-Computer science and Engineering

Dear Editors and Reviewers

Thank you so much for allowing revision for our paper. We have given the response to the reviewer’s comments and suggestions below. Besides, we also revised everything suggested carefully in the paper. We believe that the paper is ready for publication now. Thank you so much for your assistance.

Sincerely,

Authors

Brazil-India

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The author has made the necessary changes.