Deep-learning based detection of eosinophilic esophagitis. Guimarães, P., Keller, A., Fehlmann, T., Lammert, F., & Casper, M. Endoscopy, 05, 2021. ![Deep-learning based detection of eosinophilic esophagitis [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper doi abstract bibtex 4 downloads Background and aims: For eosinophilic esophagitis (EoE) a substantial diagnostic delay is still a clinically relevant phenomenon. Deep learning-based algorithms have demonstrated potential in medical image analysis. Here we establish a convolutional neuronal network (CNN)-based approach that can distinguish EoE from normal findings and candida esophagitis. Methods: We trained and tested a CNN using 484 real-world endoscopic images from 134 subjects consisting of three classes (normal, EoE, and candidiasis). Images were split into two completely independent datasets. The proposed approach was evaluated against three trainee endoscopists on the test set. Model-explainability was enhanced by deep Taylor decomposition. Results: Global accuracy (0.915 [0.880-0.940]), sensitivity (0.871 [0.819-0.910]) and specificity (0.936 [0.910-0.955]) were significantly higher than for endoscopists on the test set. Global area under the ROC curve was 0.966 [0.954-0.975]. Results were highly reproducible. Explainability analysis found that the algorithm identified characteristic signs also used by endoscopists. Conclusions: Complex endoscopic classification tasks including more than two classes can be solved by CNN-based algorithms. Thus, our algorithm (https://ccb-test.cs.uni-saarland.de/EoE/) may assist clinicians in making the diagnosis of EoE.
@article{gp0521a,
author = {Guimarães, Pedro and Keller, Andreas and Fehlmann, Tobias and Lammert, Frank and Casper, Mary},
year = {2021},
month = {05},
pages = {},
abstract = {Background and aims: For eosinophilic esophagitis (EoE) a substantial diagnostic delay is still a clinically relevant phenomenon. Deep learning-based algorithms have demonstrated potential in medical image analysis. Here we establish a convolutional neuronal network (CNN)-based approach that can distinguish EoE from normal findings and candida esophagitis. Methods: We trained and tested a CNN using 484 real-world endoscopic images from 134 subjects consisting of three classes (normal, EoE, and candidiasis). Images were split into two completely independent datasets. The proposed approach was evaluated against three trainee endoscopists on the test set. Model-explainability was enhanced by deep Taylor decomposition. Results: Global accuracy (0.915 [0.880-0.940]), sensitivity (0.871 [0.819-0.910]) and specificity (0.936 [0.910-0.955]) were significantly higher than for endoscopists on the test set. Global area under the ROC curve was 0.966 [0.954-0.975]. Results were highly reproducible. Explainability analysis found that the algorithm identified characteristic signs also used by endoscopists. Conclusions: Complex endoscopic classification tasks including more than two classes can be solved by CNN-based algorithms. Thus, our algorithm (https://ccb-test.cs.uni-saarland.de/EoE/) may assist clinicians in making the diagnosis of EoE. },
title = {Deep-learning based detection of eosinophilic esophagitis},
journal = {Endoscopy},
doi = {10.1055/a-1520-8116},
URL = {https://www.thieme-connect.de/products/ejournals/abstract/10.1055/a-1520-8116}
}
Downloads: 4
{"_id":"irRgQNCxrdtL5paPr","bibbaseid":"guimares-keller-fehlmann-lammert-casper-deeplearningbaseddetectionofeosinophilicesophagitis-2021","author_short":["Guimarães, P.","Keller, A.","Fehlmann, T.","Lammert, F.","Casper, M."],"bibdata":{"bibtype":"article","type":"article","author":[{"propositions":[],"lastnames":["Guimarães"],"firstnames":["Pedro"],"suffixes":[]},{"propositions":[],"lastnames":["Keller"],"firstnames":["Andreas"],"suffixes":[]},{"propositions":[],"lastnames":["Fehlmann"],"firstnames":["Tobias"],"suffixes":[]},{"propositions":[],"lastnames":["Lammert"],"firstnames":["Frank"],"suffixes":[]},{"propositions":[],"lastnames":["Casper"],"firstnames":["Mary"],"suffixes":[]}],"year":"2021","month":"05","pages":"","abstract":"Background and aims: For eosinophilic esophagitis (EoE) a substantial diagnostic delay is still a clinically relevant phenomenon. Deep learning-based algorithms have demonstrated potential in medical image analysis. Here we establish a convolutional neuronal network (CNN)-based approach that can distinguish EoE from normal findings and candida esophagitis. Methods: We trained and tested a CNN using 484 real-world endoscopic images from 134 subjects consisting of three classes (normal, EoE, and candidiasis). Images were split into two completely independent datasets. The proposed approach was evaluated against three trainee endoscopists on the test set. Model-explainability was enhanced by deep Taylor decomposition. Results: Global accuracy (0.915 [0.880-0.940]), sensitivity (0.871 [0.819-0.910]) and specificity (0.936 [0.910-0.955]) were significantly higher than for endoscopists on the test set. Global area under the ROC curve was 0.966 [0.954-0.975]. Results were highly reproducible. Explainability analysis found that the algorithm identified characteristic signs also used by endoscopists. Conclusions: Complex endoscopic classification tasks including more than two classes can be solved by CNN-based algorithms. Thus, our algorithm (https://ccb-test.cs.uni-saarland.de/EoE/) may assist clinicians in making the diagnosis of EoE. ","title":"Deep-learning based detection of eosinophilic esophagitis","journal":"Endoscopy","doi":"10.1055/a-1520-8116","url":"https://www.thieme-connect.de/products/ejournals/abstract/10.1055/a-1520-8116","bibtex":"@article{gp0521a,\n author = {Guimarães, Pedro and Keller, Andreas and Fehlmann, Tobias and Lammert, Frank and Casper, Mary},\n year = {2021},\n month = {05},\n pages = {},\n abstract = {Background and aims: For eosinophilic esophagitis (EoE) a substantial diagnostic delay is still a clinically relevant phenomenon. Deep learning-based algorithms have demonstrated potential in medical image analysis. Here we establish a convolutional neuronal network (CNN)-based approach that can distinguish EoE from normal findings and candida esophagitis. Methods: We trained and tested a CNN using 484 real-world endoscopic images from 134 subjects consisting of three classes (normal, EoE, and candidiasis). Images were split into two completely independent datasets. The proposed approach was evaluated against three trainee endoscopists on the test set. Model-explainability was enhanced by deep Taylor decomposition. Results: Global accuracy (0.915 [0.880-0.940]), sensitivity (0.871 [0.819-0.910]) and specificity (0.936 [0.910-0.955]) were significantly higher than for endoscopists on the test set. Global area under the ROC curve was 0.966 [0.954-0.975]. Results were highly reproducible. Explainability analysis found that the algorithm identified characteristic signs also used by endoscopists. Conclusions: Complex endoscopic classification tasks including more than two classes can be solved by CNN-based algorithms. Thus, our algorithm (https://ccb-test.cs.uni-saarland.de/EoE/) may assist clinicians in making the diagnosis of EoE. },\n title = {Deep-learning based detection of eosinophilic esophagitis},\n journal = {Endoscopy},\n doi = {10.1055/a-1520-8116},\n URL = {https://www.thieme-connect.de/products/ejournals/abstract/10.1055/a-1520-8116}\n}\n","author_short":["Guimarães, P.","Keller, A.","Fehlmann, T.","Lammert, F.","Casper, M."],"key":"gp0521a","id":"gp0521a","bibbaseid":"guimares-keller-fehlmann-lammert-casper-deeplearningbaseddetectionofeosinophilicesophagitis-2021","role":"author","urls":{"Paper":"https://www.thieme-connect.de/products/ejournals/abstract/10.1055/a-1520-8116"},"metadata":{"authorlinks":{}},"downloads":4,"html":""},"bibtype":"article","biburl":"https://www.ccb.uni-saarland.de/wp-content/uploads/2024/11/references.bib_.txt","dataSources":["iQsmnqgonvyW7tRge","RjjDBMYeiCRMZWAvn","pTW7v7XACewjrTXET","BD2qbudjMvyXtTiz5","NmhXQcJvRc2QhnSZF","ipvH6pWABxuwdKDLx","Pny5E4E9kc7C8gG8g","SiGP46KPWizw6ihLJ","ZKiRa4gncFJ5e6f9M","CZZSbiMkXJgDMN2Ei","fMYw4bZ8PtmEvvgdF","XiRWyepSYzzAnCRoW","nqMohMYmMdCvacEct"],"keywords":[],"search_terms":["deep","learning","based","detection","eosinophilic","esophagitis","guimarães","keller","fehlmann","lammert","casper"],"title":"Deep-learning based detection of eosinophilic esophagitis","year":2021,"downloads":4}