Next Article in Journal
Implementing a Wide-Area Network and Low Power Solution Using Long-Range Wide-Area Network Technology
Next Article in Special Issue
AI Diffusion Model-Based Technology for Automating the Multi-Class Labeling of Electron Microscopy Datasets of Brain Cell Organelles for Their Augmentation and Synthetic Generation
Previous Article in Journal
Application of Graph Theory and Variants of Greedy Graph Coloring Algorithms for Optimization of Distributed Peer-to-Peer Blockchain Networks
Previous Article in Special Issue
Real-Time Deployment of Ultrasound Image Interpretation AI Models for Emergency Medicine Triage Using a Swine Model
 
 
Article
Peer-Review Record

Towards Transparent AI in Medicine: ECG-Based Arrhythmia Detection with Explainable Deep Learning

Technologies 2025, 13(1), 34; https://doi.org/10.3390/technologies13010034
by Oleksii Kovalchuk 1, Oleksandr Barmak 1, Pavlo Radiuk 1,*, Liliana Klymenko 2 and Iurii Krak 3,4
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Reviewer 3: Anonymous
Technologies 2025, 13(1), 34; https://doi.org/10.3390/technologies13010034
Submission received: 8 November 2024 / Revised: 5 January 2025 / Accepted: 11 January 2025 / Published: 14 January 2025

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The authors present the article entitled “Towards Transparent AI in Medicine: ECG-Based Arrhythmia Detection with Explainable Deep Learning”. In general, the manuscript is well-structured and is easy to follow. The authors provide a comparison of the proposed method vs the already reported in the literature by highlighting the main findings and contributions. The manuscript presents the following minor concerns:

Figure 9: Please add x and y axis titles.

Figure 16: Please explain why the model did not classified class 2.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Congratulations on this extensive and well-executed work. The results achieved with the data sets used, compared to those published elsewhere, are very impressive.

Here are a few suggestions for improvements from the reviewer's point of view:

Line 193: the ECG is not the course of the amplitude over time, but rather the course of the time signal (amplitude is the wrong term in this sentence).

Line 817: first described limitation, the proposed approach is its dependence on the accuracy: perhaps the inclusion of historical data for the RR intervals could improve the R peak detection (since these always depend on physiological factors, such as the increasing or decreasing physical stress on the patient), provided that in individual cases the her rate variability does not predominate. This circumstance should perhaps be included in the discussion.

Line 824; second described limitation of the proposed approach: this limitation could be counteracted by including another data set, which in general shows these very rarely occurring pathological anomalies. For this purpose, it would be of interest to specifically generate such a data set of patients with these symptoms. This aspect should also be included in the discussion.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript "Towards Transparent AI in Medicine: ECG-Based Arrhythmia Detection with Explainable Deep Learning" focuses on the use of artificial intelligence in the interpretation of cardiograms. This is very important and necessary work that will allow physicians to quickly and efficiently identify cardiovascular pathology. The work is based on a shared principle, everything is clear and understandable. Using this MIT-BIH 24 database, the developed approach achieved an accuracy of 99.43% and F1 scores approaching 100% for the main 25 classes of arrhythmias. There are no fundamental comments on the work. The only thing we would like to see is the mention of some clinical cases, at least in a file with additional information that would allow us to demonstrate the developed algorithm in action. 

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Back to TopTop