Next Article in Journal
Traditional Grain-Based vs. Commercial Milk Kefirs, How Different Are They?
Next Article in Special Issue
Obfuscation Algorithm for Privacy-Preserving Deep Learning-Based Medical Image Analysis
Previous Article in Journal
Creating a Novel Mathematical Model of the Kv10.1 Ion Channel and Controlling Channel Activity with Nanoelectromechanical Systems
Previous Article in Special Issue
Real-Time Multi-Label Upper Gastrointestinal Anatomy Recognition from Gastroscope Videos
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Applied Sciences
Volume 12
Issue 8
10.3390/app12083839
Review Report
applsci-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
Article
Peer-Review Record

Normalizing Flows for Out-of-Distribution Detection: Application to Coronary Artery Segmentation

Appl. Sci. 2022, 12(8), 3839; https://doi.org/10.3390/app12083839
by Costin Florian Ciușdel 1,2,*, Lucian Mihai Itu 1,2, Serkan Cimen 3, Michael Wels 4, Chris Schwemmer 4, Philipp Fortner 5, Sebastian Seitz 5, Florian Andre 5, Sebastian Johannes Buß 5, Puneet Sharma 3 and Saikiran Rapaka 3
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Reviewer 3: Anonymous
Appl. Sci. 2022, 12(8), 3839; https://doi.org/10.3390/app12083839
Submission received: 28 February 2022 / Revised: 5 April 2022 / Accepted: 6 April 2022 / Published: 11 April 2022
(This article belongs to the Special Issue Deep Neural Networks in Medical Imaging)

Round 1

Reviewer 1 Report

Dear authors, 

After reading your manuscript I got some unresolved questions, so I would like to ask you about them:

First, what are the semantic features do you point out?

Second, why do you compare your model against base model on generated samples? I mean, knowing it did not produce a segmentation mask.

Third, how much time did you reduce using this approach vs the actual approach.

Thanks in advance for your consideration and time to respond these inquiries.

Best regards,   

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

Reviewing the manuscript entitled, “Normalizing Flows for Out-of-Distribution Detection: Application to Coronary Artery Segmentation” by Ciusdel CF et al., this is an article focusing on establishment of CCTA with higher diagnostic accuracy. This is a very important manuscript, and when only less invasive CCTA can make a diagnosis of coronary artery disease, the contribution to medical care will be immeasurable. However, it is a rather esoteric manuscript for the cardiologist who will be the main leader. Therefore, the authors need to respond to the following concerns.

In figure 6 to 12, the authors should describe how to interpret the difference between baseline and proposed architecture of the AuRoC performance for the cardiologist to understand easily.

Accurate diagnosis of coronary artery lesions must lead to optimization of treatment, and if correct images cannot be obtained, coronary angiography is all that is needed. The author should describe the relationship between higher image detection of CCTA that you proposed and coronary angiography as a benefit to the patient who has coronary heart disease in the discussion.

The authors need to add the abbreviation table. 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

Here authors nicely evaluated an approach based on normalizing flows architectures for the Out-of-Distribution detection of coronary lumen segmentations by using a new type of coupling layer.

All controls has been tested, approach is very well described and conclusions are supported by data.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Back to TopTop