Multi-encoder U-Net for oral squamous cell carcinoma image segmentation
Abstract: Oral tumors are responsible for about 170,000 deaths every year in the World. In this paper, we focus on oral squamous cell carcinoma (OSCC), which represents up to 80-90% of all malignant neoplasms of the oral cavity. We present a novel deep learning-based method for segmenting whole slid...
Published in: | 2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA) |
---|---|
Main Authors: | , , , , |
Format: | Conference Object |
Language: | English |
Published: |
2022
|
Subjects: | |
Online Access: | https://hdl.handle.net/10067/1915630151162165141 |
id |
ftunivantwerpen:c:irua:191563 |
---|---|
record_format |
openpolar |
spelling |
ftunivantwerpen:c:irua:191563 2024-09-09T20:02:22+00:00 Multi-encoder U-Net for oral squamous cell carcinoma image segmentation Pennisi, Andrea Bloisi, Domenico D. Nardi, Daniele Varricchio, Silvia Merolla, Francesco 2022 https://hdl.handle.net/10067/1915630151162165141 eng eng info:eu-repo/semantics/altIdentifier/doi/10.1109/MEMEA54994.2022.9856482 info:eu-repo/semantics/altIdentifier/isi/000861225100076 info:eu-repo/semantics/closedAccess 17th IEEE International Symposium on Medical Measurements and, Applications (IEEE MeMeA), JUN 22-24, 2022, Messina, Italy 978-1-6654-8299-8 2022 IEEE International Symposium on Medical Measurements and Applications (MEMEA 2022) Computer. Automation info:eu-repo/semantics/conferenceObject 2022 ftunivantwerpen https://doi.org/10.1109/MEMEA54994.2022.9856482 2024-06-18T14:20:24Z Abstract: Oral tumors are responsible for about 170,000 deaths every year in the World. In this paper, we focus on oral squamous cell carcinoma (OSCC), which represents up to 80-90% of all malignant neoplasms of the oral cavity. We present a novel deep learning-based method for segmenting whole slide image (WSI) samples at the pixel level. The proposed method is a modification of the well-known U-Net architecture through a multi-encoder structure. In particular, our network, called Multi encoder U-Net, is a multi-encoder single decoder network that takes as input an image and splits it in tiles. For each tile, there is an encoder responsible for encoding it in the latent space, then a convolutional layer is responsible for merging the tiles into a single layer. Each layer of the decoder takes as input the previous up-sampled layer and concatenate it with the layer made by merging the corresponding layers of the multiple encoders. Experiments have been carried out on the publicly available ORal Cancer Annotated (ORCA) dataset, which contains annotated data from the TCGA repository. Quantitative experimental results, obtained using three different quality metrics, demonstrate the effectiveness of the proposed approach, which achieves 82% Pixel-wise Accuracy, 0.82 Dice similarity score, and 0.72 Mean Intersection Over Union. Conference Object Orca IRUA - Institutional Repository van de Universiteit Antwerpen 2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA) 1 6 |
institution |
Open Polar |
collection |
IRUA - Institutional Repository van de Universiteit Antwerpen |
op_collection_id |
ftunivantwerpen |
language |
English |
topic |
Computer. Automation |
spellingShingle |
Computer. Automation Pennisi, Andrea Bloisi, Domenico D. Nardi, Daniele Varricchio, Silvia Merolla, Francesco Multi-encoder U-Net for oral squamous cell carcinoma image segmentation |
topic_facet |
Computer. Automation |
description |
Abstract: Oral tumors are responsible for about 170,000 deaths every year in the World. In this paper, we focus on oral squamous cell carcinoma (OSCC), which represents up to 80-90% of all malignant neoplasms of the oral cavity. We present a novel deep learning-based method for segmenting whole slide image (WSI) samples at the pixel level. The proposed method is a modification of the well-known U-Net architecture through a multi-encoder structure. In particular, our network, called Multi encoder U-Net, is a multi-encoder single decoder network that takes as input an image and splits it in tiles. For each tile, there is an encoder responsible for encoding it in the latent space, then a convolutional layer is responsible for merging the tiles into a single layer. Each layer of the decoder takes as input the previous up-sampled layer and concatenate it with the layer made by merging the corresponding layers of the multiple encoders. Experiments have been carried out on the publicly available ORal Cancer Annotated (ORCA) dataset, which contains annotated data from the TCGA repository. Quantitative experimental results, obtained using three different quality metrics, demonstrate the effectiveness of the proposed approach, which achieves 82% Pixel-wise Accuracy, 0.82 Dice similarity score, and 0.72 Mean Intersection Over Union. |
format |
Conference Object |
author |
Pennisi, Andrea Bloisi, Domenico D. Nardi, Daniele Varricchio, Silvia Merolla, Francesco |
author_facet |
Pennisi, Andrea Bloisi, Domenico D. Nardi, Daniele Varricchio, Silvia Merolla, Francesco |
author_sort |
Pennisi, Andrea |
title |
Multi-encoder U-Net for oral squamous cell carcinoma image segmentation |
title_short |
Multi-encoder U-Net for oral squamous cell carcinoma image segmentation |
title_full |
Multi-encoder U-Net for oral squamous cell carcinoma image segmentation |
title_fullStr |
Multi-encoder U-Net for oral squamous cell carcinoma image segmentation |
title_full_unstemmed |
Multi-encoder U-Net for oral squamous cell carcinoma image segmentation |
title_sort |
multi-encoder u-net for oral squamous cell carcinoma image segmentation |
publishDate |
2022 |
url |
https://hdl.handle.net/10067/1915630151162165141 |
genre |
Orca |
genre_facet |
Orca |
op_source |
17th IEEE International Symposium on Medical Measurements and, Applications (IEEE MeMeA), JUN 22-24, 2022, Messina, Italy 978-1-6654-8299-8 2022 IEEE International Symposium on Medical Measurements and Applications (MEMEA 2022) |
op_relation |
info:eu-repo/semantics/altIdentifier/doi/10.1109/MEMEA54994.2022.9856482 info:eu-repo/semantics/altIdentifier/isi/000861225100076 |
op_rights |
info:eu-repo/semantics/closedAccess |
op_doi |
https://doi.org/10.1109/MEMEA54994.2022.9856482 |
container_title |
2022 IEEE International Symposium on Medical Measurements and Applications (MeMeA) |
container_start_page |
1 |
op_container_end_page |
6 |
_version_ |
1809934327836835840 |