Reduction of Video Capsule Endoscopy Reading Times Using Deep Learning with Small Data

Round 1

Reviewer 1 Report (Previous Reviewer 2)

The following reference should be added, to balance the report from a certain small bowel manufacturer:

Xie X, Xiao YF, Zhao XY, Li JJ, Yang QQ, Peng X, Nie XB, Zhou JY, Zhao YB, Yang H, Liu X, Liu E, Chen YY, Zhou YY, Fan CQ, Bai JY, Lin H, Koulaouzidis A, Yang SM. Development and Validation of an Artificial Intelligence Model for Small Bowel Capsule Endoscopy Video Review. JAMA Netw Open. 2022 Jul 1;5(7):e2221992. doi: 10.1001/jamanetworkopen.2022.21992

Furthermore, the following should be added: Dray X, Iakovidis D, Houdeville C, Jover R, Diamantis D, Histace A, Koulaouzidis A. Artificial intelligence in small bowel capsule endoscopy - current status, challenges and future promise. J Gastroenterol Hepatol. 2021 Jan;36(1):12-19. doi: 10.1111/jgh.15341

Author Response

The following reference should be added, to balance the report from a certain small bowel manufacturer:

Xie X, Xiao YF, Zhao XY, Li JJ, Yang QQ, Peng X, Nie XB, Zhou JY, Zhao YB, Yang H, Liu X, Liu E, Chen YY, Zhou YY, Fan CQ, Bai JY, Lin H, Koulaouzidis A, Yang SM. Development and Validation of an Artificial Intelligence Model for Small Bowel Capsule Endoscopy Video Review. JAMA Netw Open. 2022 Jul 1;5(7):e2221992. doi:10.1001/jamanetworkopen.2022.21992

This citation was added to the manuscript in the discussion and can be found on lines 179 to 181. 

Furthermore, the following should be added:

Dray X, Iakovidis D, Houdeville C, Jover R, Diamantis D, Histace A, Koulaouzidis A. Artificial intelligence in small bowel capsule endoscopy - current status, challenges and future promise. J Gastroenterol Hepatol. 2021 Jan;36(1):12-19. doi: 10.1111/jgh.15341

This citation was added to the manuscript in the introduction and can be found on line 31.

Reviewer 2 Report (New Reviewer)

In this paper Morera et.al. have explored CNN to reduce reading times of video capsule endoscopy. Overall, this is a well researched paper and its a significant scientific importance. I wish the presentation is more vivid next time. The authors have already acknowledged the limitations of this study which are spot on. Only criticism for future publications is that the paper should have more figures and information. For example, they did t-test but I could not find that information in whole manuscript. It does not walks the audience through the information. 

Overall, this manuscript is ready for publication. Congratulations!

Author Response

In this paper Morera et.al. have explored CNN to reduce reading times of video capsule endoscopy. Overall, this is a well-researched paper and it’s a significant scientific importance. I wish the presentation is more vivid next time. The authors have already acknowledged the limitations of this study which are spot on. Only criticism for future publications is that the paper should have more figures and information. For example, they did t-test but I could not find that information in whole manuscript. It does not walk the audience through the information. Overall, this manuscript is ready for publication. Congratulations!

Thank you for the kind words. We have re-evaluated the paired t-test and provided more in-depth information for both tests done. The first t-test comparing a single network vs.  ensemble without augmentation can be found in the manuscript in Table 2. The second t-test comparing the ensemble without augmentation vs. ensemble with augmentation can be found in the manuscript in Table 5.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

In 2022 we have already a large literature in this field, including multi brand CNN, with several algorithms reported, tested in hundreds of videos.

Despite the efforts of the authors, a model that was trained in only two videos and tested in one does not provide new information that is scientifically sound.

Reviewer 2 Report

Please check a few minor grammar language issues, e.g., missing articles etc.

Please present your methods in a more palatable manner, i.e., if a big chunk of one paragraph is not helpful, break it into 2-3 at least.

The references should be updated; some of them include only the first authors .etc., and up-to-date references in the introduction should include:

Beg S, Card T, Sidhu R, Wronska E, Ragunath K; UK capsule endoscopy users' group. The impact of reader fatigue on the accuracy of capsule endoscopy interpretation. Dig Liver Dis. 2021 Aug;53(8):1028-1033. DOI: 10.1016/j.dld.2021.04.024

Dray X, Toth E, de Lange T, Koulaouzidis A. Artificial intelligence, capsule endoscopy, databases, and the Sword of Damocles. Endosc Int Open. 2021 Nov 12;9(11):E1754-E1755. DOI: 10.1055/a-1521-4882.

AND

Rondonotti E, Pennazio M, Toth E, Koulaouzidis A. How to read small bowel capsule endoscopy: a practical guide for everyday use. Endosc Int Open. 2020 Oct;8(10):E1220-E1224. DOI: 10.1055/a-1210-4830.

I also recommend using the 'κάψουλα' instead of the current 'káψoυλa' in line 37.

Moreover, please remove the 'maintained by the university of Thessaly' and only keep the publicly available database from your abstract and methods.

The discussion is short and snappy.