Comparative analysis of automatic segmentation of esophageal cancer using 3D Res-UNet on conventional and 40-keV virtual mono-energetic CT Images: a retrospective study

To assess the performance of 3D Res-UNet for fully automated segmentation of esophageal cancer (EC) and compare the segmentation accuracy between conventional images (CI) and 40-keV virtual mono-energetic images (VMI ). Patients underwent spectral CT scanning and diagnosed of EC by operation or gast...

Full description

Saved in:
Bibliographic Details
Published in:PeerJ (San Francisco, CA) CA), 2023-07, Vol.11, p.e15707-e15707, Article e15707
Main Authors: Zhong, Hua, Li, Anqi, Chen, Yingdong, Huang, Qianwen, Chen, Xingbiao, Kang, Jianghe, You, Youkuang
Format: Article
Language:English
Subjects:
3D Res-UNet
Automatic segmentation
Comparative analysis
Computational Biology
CT imaging
Diagnostic imaging
Dual-layer spectral-detector computed tomography
Esophageal cancer
Gastroenterology and Hepatology
Medical imaging equipment
Neural networks
Radiology and Medical Imaging
Virtual Monoenergetic Imaging
Citations: Items that this one cites
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:To assess the performance of 3D Res-UNet for fully automated segmentation of esophageal cancer (EC) and compare the segmentation accuracy between conventional images (CI) and 40-keV virtual mono-energetic images (VMI ). Patients underwent spectral CT scanning and diagnosed of EC by operation or gastroscope biopsy in our hospital from 2019 to 2020 were analyzed retrospectively. All artery spectral base images were transferred to the dedicated workstation to generate VMI and CI. The segmentation model of EC was constructed by 3D Res-UNet neural network in VMI and CI, respectively. After optimization training, the Dice similarity coefficient (DSC), overlap (IOU), average symmetrical surface distance (ASSD) and 95% Hausdorff distance (HD_95) of EC at pixel level were tested and calculated in the test set. The paired rank sum test was used to compare the results of VMI and CI. A total of 160 patients were included in the analysis and randomly divided into the training dataset (104 patients), validation dataset (26 patients) and test dataset (30 patients). VMI as input data in the training dataset resulted in higher model performance in the test dataset in comparison with using CI as input data (DSC:0.875 0.859, IOU: 0.777 0.755, ASSD:0.911 0.981, HD_95: 4.41 6.23, all -value
ISSN:2167-8359
2167-8359
DOI:10.7717/peerj.15707