A Deep Learning Workflow for Mass-Forming Intrahepatic Cholangiocarcinoma and Hepatocellular Carcinoma Classification Based on MRI
Round 1
Reviewer 1 Report
Overall the presented work is significant and addresses the gap in existing research. I have a minor concern that could be modified:
1) The tumor lesion boundary denoted by the radiologists involves a significant bias which can be improved by introducing an automatized step. If the authors wish to keep the radiologist's contribution intact in the work, then details should be provided on how many radiologists drew the lesion boundaries, and what were the inter-reader/intra-reader variability. There might be a certain bias in the lesion drawing pattern (where they already know if it is an MF_ICC or HCC) which could be discussed more.
Author Response
Please see the attachment
Author Response File: 
Author Response.docx
Reviewer 2 Report
The article is interesting and well written. The authors precisely explained its purpose and assumptions. I have some comments on the content of the article. I indicate them below.
1. One paragraph can be added in the introduction section about the flow of the paper organization.
2. In the subsection describing the material for research and its division into classes, it is worth adding sample images representing individual classes of images.
3. After the representation of the results, it would be necessary to add examples of images that were incorrectly classified.
4. The discussion part should include a comparison of the authors' results with other research results presented in the literature.
5. Future directions can be added in the new paragraph.
6. Literature is limited. It is worth adding a few items describing the latest research.
Author Response
Please see the attachment
Author Response File: Author Response.docx
Reviewer 3 Report
1. Title should be written “Capitalized” and shortened.
2. The language is good. However, it needs some proofreading
3. Keywords: only the first term should be written “Capitalized”. The next terms should be written in lower cases.
4. Abstract is good, but can be improved. Use the main motivation and contributions of the work. It should answer the questions: What problem did you study and why is it important? What methods did you use? What were your main results? And what conclusions can you draw from your results?
5. It is advisable to divide the introduction into three subsections: 1) Motivation, 2) Literature review, and 3) contributions.
6. Please, include the following references in the related part of your work: (2018). MRI features of combined hepatocellular-cholangiocarcinoma versus mass forming intrahepatic cholangiocarcinoma. Cancer Imaging, 18(1), 1-9; (2021). A new optimized sequential method for lung tumor diagnosis based on deep learning and converged search and rescue algorithm. Biomedical Signal Processing and Control, 68, 102761; (2022). Added-value of ancillary imaging features for differentiating hepatocellular carcinoma from intrahepatic mass-forming cholangiocarcinoma on Gd-BOPTA-enhanced MRI in LI-RADS M. Abdominal Radiology, 47(3), 957-968; (2021). Novel computer‐aided lung cancer detection based on convolutional neural network‐based and feature‐based classifiers using metaheuristics. International Journal of Imaging Systems and Technology, 31(4), 1954-1969.
7. Please provide full details of the advantages of your work.
8. The disadvantages of the existing methods are missing.
9. Conclusion should be rewritten focusing on the main outputs of the paper with some important numerical results of the method. Also, conclusions section should be considered as a separated part from the paper, so, all of the work should be pointed in it.
Author Response
Please see the attachment
Author Response File: Author Response.docx
Round 2
Reviewer 2 Report
I accept the revised manuscript.
Reviewer 3 Report
The authors resolved all of my concerns.
