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.
The paper has been revised and improved. As a result, the paper is ready to be accepted.
Two reviewers have consistent recommendations. As a result, a revision is suggested.
[# PeerJ Staff Note: Please ensure that all review comments are addressed in a rebuttal letter and any edits or clarifications mentioned in the letter are also inserted into the revised manuscript where appropriate. It is a common mistake to address reviewer questions in the rebuttal letter but not in the revised manuscript. If a reviewer raised a question then your readers will probably have the same question so you should ensure that the manuscript can stand alone without the rebuttal letter. Directions on how to prepare a rebuttal letter can be found at: https://peerj.com/benefits/academic-rebuttal-letters/ #]
In this paper a novel convolutional neural network is proposed for determination of abnormalities or infections in human tympanic membrane using image processing methods. The paper technically seems useful regarding experimental applications. The proposed study aims to classify middle ear otoscope images based on the deep convolution neural networks (CNNs). To this aim, a novel deep CNN model including a combination of the channel and spatial model (CBAM), residual blocks, and hypercolumn technique have been embedded into the proposed model. The achieved results have been evaluated as promising.
To improve the quality of study, I have presented my comment as follows:
1. According to my knowledge, earwax is not an OM type, please clarity this issue on the manuscript.
2. There is a need for a more detailed explanation of some of the details of the Eardrum database in particular.
3. Authors must explain clearly the advantage of proposed algorithm.
4. In related work, paper with reference number [19] author mentioned the number of image sample as 10.544. How is it possible?
5. What is the initial learning rate vale of proposed model for training the dataset?
6. How many Positive and negative images taken for performing the tests of proposed model?
7. Author mentioned that training period shortened for the proposed work when compared with pre-trained CNNs, as we know transfer-learning takes less amount of time when compared to learning from basic on the particular dataset.
After minor revision, I think the manuscript can be proposed for publishing.
With regards.
The authors use methods that seem to be familiar to those in computer science, which I am not. Similarly, I am not the intended audience and could not understand the terminology.
As an example: the "down" function is defined in terms of the "down" function and it was not any clearer after reading the material. So I assume most readers are already familiar with this.
The emphasis on computational efficiency is important but does not appear useful in a data set with only few hundred observations. Otherwise this is OK.
They compare several clustering methods. This is appropriate.
The findings appear valid.
What is the "gold standard" for evaluating sensitivity and specificity? The article begins by saying the visual observation is "subjective and error-prone" so there is no way of telling whether their methods are correctly classifying OM cases.
Nothing additional
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.