Detect-and-segment: A deep learning approach to automate wound image segmentation. Scebba, G., Zhang, J., Catanzaro, S., Mihai, C., Distler, O., Berli, M., & Karlen, W. Informatics in Medicine Unlocked, 29:100884, 2022. doi abstract bibtex Chronic wounds significantly impact quality of life. They can rapidly deteriorate and require close monitoring of healing progress. Image-based wound analysis is a way of objectively assessing the wound status by quantifying important features that are related to healing. However, high heterogeneity of the wound types and imaging conditions challenge the robust segmentation of wound images. We present Detect-and-Segment (DS), a deep learning approach to produce wound segmentation maps with high generalization capabilities. In our approach, dedicated deep neural networks detected the wound position, isolated the wound from the perturbing background, and computed a wound segmentation map. We tested this approach on a diabetic foot ulcers data set and compared it to a segmentation method based on the full image. To evaluate its generalizability on out-of-distribution data, we measured the performance of the DS approach on 4 additional independent data sets, with larger variety of wound types from different body locations. The Matthews’ correlation coefficient (MCC) improved from 0.29 (full image) to 0.85 (DS) on the diabetic foot ulcer data set. When the DS was tested on the independent data sets, the mean MCC increased from 0.17 to 0.85 . Furthermore, the DS enabled the training of segmentation models with up to 90% less training data without impacting the segmentation performance. The proposed DS approach is a step towards automating wound analysis and reducing efforts to manage chronic wounds.
@article{
title = {Detect-and-segment: A deep learning approach to automate wound image segmentation},
type = {article},
year = {2022},
keywords = {Chronic wounds,Generalizability,Machine learning,Semantic segmentation,Smartphone},
pages = {100884},
volume = {29},
id = {12dafe98-05a4-318d-b4ec-735d0b9f69d4},
created = {2022-09-04T18:02:17.480Z},
file_attached = {false},
profile_id = {6d353feb-efe4-367e-84a2-0815eb9ca878},
last_modified = {2022-09-04T18:12:04.568Z},
read = {false},
starred = {false},
authored = {true},
confirmed = {true},
hidden = {false},
citation_key = {SCEBBA2022100884},
source_type = {article},
folder_uuids = {f1f67efc-95a7-4f1a-b181-c3670c667a34},
private_publication = {false},
abstract = {Chronic wounds significantly impact quality of life. They can rapidly deteriorate and require close monitoring of healing progress. Image-based wound analysis is a way of objectively assessing the wound status by quantifying important features that are related to healing. However, high heterogeneity of the wound types and imaging conditions challenge the robust segmentation of wound images. We present Detect-and-Segment (DS), a deep learning approach to produce wound segmentation maps with high generalization capabilities. In our approach, dedicated deep neural networks detected the wound position, isolated the wound from the perturbing background, and computed a wound segmentation map. We tested this approach on a diabetic foot ulcers data set and compared it to a segmentation method based on the full image. To evaluate its generalizability on out-of-distribution data, we measured the performance of the DS approach on 4 additional independent data sets, with larger variety of wound types from different body locations. The Matthews’ correlation coefficient (MCC) improved from 0.29 (full image) to 0.85 (DS) on the diabetic foot ulcer data set. When the DS was tested on the independent data sets, the mean MCC increased from 0.17 to 0.85 . Furthermore, the DS enabled the training of segmentation models with up to 90% less training data without impacting the segmentation performance. The proposed DS approach is a step towards automating wound analysis and reducing efforts to manage chronic wounds.},
bibtype = {article},
author = {Scebba, Gaetano and Zhang, Jia and Catanzaro, Sabrina and Mihai, Carina and Distler, Oliver and Berli, Martin and Karlen, Walter},
doi = {https://doi.org/10.1016/j.imu.2022.100884},
journal = {Informatics in Medicine Unlocked}
}
Downloads: 0
{"_id":"KJvkQa4q9kCTWncqv","bibbaseid":"scebba-zhang-catanzaro-mihai-distler-berli-karlen-detectandsegmentadeeplearningapproachtoautomatewoundimagesegmentation-2022","author_short":["Scebba, G.","Zhang, J.","Catanzaro, S.","Mihai, C.","Distler, O.","Berli, M.","Karlen, W."],"bibdata":{"title":"Detect-and-segment: A deep learning approach to automate wound image segmentation","type":"article","year":"2022","keywords":"Chronic wounds,Generalizability,Machine learning,Semantic segmentation,Smartphone","pages":"100884","volume":"29","id":"12dafe98-05a4-318d-b4ec-735d0b9f69d4","created":"2022-09-04T18:02:17.480Z","file_attached":false,"profile_id":"6d353feb-efe4-367e-84a2-0815eb9ca878","last_modified":"2022-09-04T18:12:04.568Z","read":false,"starred":false,"authored":"true","confirmed":"true","hidden":false,"citation_key":"SCEBBA2022100884","source_type":"article","folder_uuids":"f1f67efc-95a7-4f1a-b181-c3670c667a34","private_publication":false,"abstract":"Chronic wounds significantly impact quality of life. They can rapidly deteriorate and require close monitoring of healing progress. Image-based wound analysis is a way of objectively assessing the wound status by quantifying important features that are related to healing. However, high heterogeneity of the wound types and imaging conditions challenge the robust segmentation of wound images. We present Detect-and-Segment (DS), a deep learning approach to produce wound segmentation maps with high generalization capabilities. In our approach, dedicated deep neural networks detected the wound position, isolated the wound from the perturbing background, and computed a wound segmentation map. We tested this approach on a diabetic foot ulcers data set and compared it to a segmentation method based on the full image. To evaluate its generalizability on out-of-distribution data, we measured the performance of the DS approach on 4 additional independent data sets, with larger variety of wound types from different body locations. The Matthews’ correlation coefficient (MCC) improved from 0.29 (full image) to 0.85 (DS) on the diabetic foot ulcer data set. When the DS was tested on the independent data sets, the mean MCC increased from 0.17 to 0.85 . Furthermore, the DS enabled the training of segmentation models with up to 90% less training data without impacting the segmentation performance. The proposed DS approach is a step towards automating wound analysis and reducing efforts to manage chronic wounds.","bibtype":"article","author":"Scebba, Gaetano and Zhang, Jia and Catanzaro, Sabrina and Mihai, Carina and Distler, Oliver and Berli, Martin and Karlen, Walter","doi":"https://doi.org/10.1016/j.imu.2022.100884","journal":"Informatics in Medicine Unlocked","bibtex":"@article{\n title = {Detect-and-segment: A deep learning approach to automate wound image segmentation},\n type = {article},\n year = {2022},\n keywords = {Chronic wounds,Generalizability,Machine learning,Semantic segmentation,Smartphone},\n pages = {100884},\n volume = {29},\n id = {12dafe98-05a4-318d-b4ec-735d0b9f69d4},\n created = {2022-09-04T18:02:17.480Z},\n file_attached = {false},\n profile_id = {6d353feb-efe4-367e-84a2-0815eb9ca878},\n last_modified = {2022-09-04T18:12:04.568Z},\n read = {false},\n starred = {false},\n authored = {true},\n confirmed = {true},\n hidden = {false},\n citation_key = {SCEBBA2022100884},\n source_type = {article},\n folder_uuids = {f1f67efc-95a7-4f1a-b181-c3670c667a34},\n private_publication = {false},\n abstract = {Chronic wounds significantly impact quality of life. They can rapidly deteriorate and require close monitoring of healing progress. Image-based wound analysis is a way of objectively assessing the wound status by quantifying important features that are related to healing. However, high heterogeneity of the wound types and imaging conditions challenge the robust segmentation of wound images. We present Detect-and-Segment (DS), a deep learning approach to produce wound segmentation maps with high generalization capabilities. In our approach, dedicated deep neural networks detected the wound position, isolated the wound from the perturbing background, and computed a wound segmentation map. We tested this approach on a diabetic foot ulcers data set and compared it to a segmentation method based on the full image. To evaluate its generalizability on out-of-distribution data, we measured the performance of the DS approach on 4 additional independent data sets, with larger variety of wound types from different body locations. The Matthews’ correlation coefficient (MCC) improved from 0.29 (full image) to 0.85 (DS) on the diabetic foot ulcer data set. When the DS was tested on the independent data sets, the mean MCC increased from 0.17 to 0.85 . Furthermore, the DS enabled the training of segmentation models with up to 90% less training data without impacting the segmentation performance. The proposed DS approach is a step towards automating wound analysis and reducing efforts to manage chronic wounds.},\n bibtype = {article},\n author = {Scebba, Gaetano and Zhang, Jia and Catanzaro, Sabrina and Mihai, Carina and Distler, Oliver and Berli, Martin and Karlen, Walter},\n doi = {https://doi.org/10.1016/j.imu.2022.100884},\n journal = {Informatics in Medicine Unlocked}\n}","author_short":["Scebba, G.","Zhang, J.","Catanzaro, S.","Mihai, C.","Distler, O.","Berli, M.","Karlen, W."],"biburl":"https://bibbase.org/service/mendeley/6d353feb-efe4-367e-84a2-0815eb9ca878","bibbaseid":"scebba-zhang-catanzaro-mihai-distler-berli-karlen-detectandsegmentadeeplearningapproachtoautomatewoundimagesegmentation-2022","role":"author","urls":{},"keyword":["Chronic wounds","Generalizability","Machine learning","Semantic segmentation","Smartphone"],"metadata":{"authorlinks":{}}},"bibtype":"article","biburl":"https://bibbase.org/service/mendeley/6d353feb-efe4-367e-84a2-0815eb9ca878","dataSources":["m8Y57GfgnRrMKZTQS","ya2CyA73rpZseyrZ8","epk5yKhDyD37NAsSC","93mP5kMGHQy5bo5xJ"],"keywords":["chronic wounds","generalizability","machine learning","semantic segmentation","smartphone"],"search_terms":["detect","segment","deep","learning","approach","automate","wound","image","segmentation","scebba","zhang","catanzaro","mihai","distler","berli","karlen"],"title":"Detect-and-segment: A deep learning approach to automate wound image segmentation","year":2022}