All reviews of published articles are made public. This includes manuscript files, peer review comments, author rebuttals and revised materials. Note: This was optional for articles submitted before 13 February 2023.
Peer reviewers are encouraged (but not required) to provide their names to the authors when submitting their peer review. If they agree to provide their name, then their personal profile page will reflect a public acknowledgment that they performed a review (even if the article is rejected). If the article is accepted, then reviewers who provided their name will be associated with the article itself.
I'm happy with accepting the revised paper.
[# PeerJ Staff Note - this decision was reviewed and approved by Xiangjie Kong, a 'PeerJ Computer Science' Section Editor covering this Section #]
The authors addressed all the comments. Overall, the paper is well-written and clear for readers to understand.
No comment
No comment
The authors are encouraged to fully justify the contributions with more experimental analysis and technical details of the datasets as the reviewers recommended.
[# PeerJ Staff Note: Please ensure that all review and editorial comments are addressed in a response letter and any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate. #]
[# PeerJ Staff Note: The review process has identified that the English language must be improved. PeerJ can provide language editing services - please contact us at copyediting@peerj.com for pricing (be sure to provide your manuscript number and title) #]
The author propose an interesting method for figure and caption extraction from medical documents. To do that, the author first apply edge detection on the scanned documents. Next, connected component are used to identify the connected regions within the document. Specific features are used to identify the graphic content and a heuristic method is used for identify caption. The method description is clear. However, the experiment only tested on two scanned books rather than test on a large number of biomedical documents. The author needs to provide more information about their experiments before the publication.
The method only tested on two scanned books. As the book typically follow the same layout for presenting figure and caption pairs, the author should test their method on a large number of biomedical publications.
Moreover, the exact number of figures, captions, and figure-caption pairs should be reported for those two books.
The evaluation metrics, precision, recall and f-score should also be further explained.
The author need either provide more details about current experiments on two books or provide more experiments on a large number of biomedical publications to prove the impact of the work.
Minor comments:
1. The section in the paper is not numbered. So, please number them first before referring to them.
2. The edge detector of Mallat needs a reference.
3. The MSER algorithm mentioned by the author cannot find in the draft.
In this paper, the authors presented a six-fold approach for extracting figures and captions from medical documents and pairing the figure extracted with the corresponding captions. The six steps of the proposed method include preprocessing, edge detection, connected component extraction, graphic objects detection, caption detection and figure-caption pair extraction.
I have the following comments for authors to consider.
The writing needs to be enhanced. Some examples where the language could be improved include lines 15, 24, 25, 56, 346. I suggest the authors have a colleague who is proficient in scientific writing review your manuscript to eliminate grammar errors and ensure the paper is comprehensive.
As introduced in Section Results (lines 316, 329, 336) , there are multiple parameters used in the presented method that are decided “after several experimentations”. What is the dataset used to tune such parameters? Are these parameters decided using the two self-created datasets that are also used to evaluate the proposed method?
The authors use two self-created databases for evaluating the proposed method and for comparing it with other previously presented methods.
Having more details of the datasets can help the readers to better understand the performance of the presented method. For instance, why were these two books chosen? How many figure-captions paper in each dataset and what is the annotation process.
There are other readily available datasets for figure-caption extraction in medical documents, such as the dataset used in citation 1. How is the performance of the presented method in such databases?
All text and materials provided via this peer-review history page are made available under a Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.