Liver segemtation in CT image with no-edge-cuting UNet

Wenbing Zhao; Kairan Li; Di Zhao; Yunpeng Jiang; Ji Wu; Jinjun Chen; Xianyue Quan; Xinming Li; Feng Xue

doi:10.1109/ITAIC49862.2020.9338815

Liver segemtation in CT image with no-edge-cuting UNet

Wenbing Zhao^*, Kairan Li, Di Zhao^*, Yunpeng Jiang, Ji Wu, Jinjun Chen, Xianyue Quan, Xinming Li, Feng Xue

^*Corresponding author for this work

Research output: Contribution to journal › Conference article › peer-review

4 Citations (Scopus)

Abstract

UNet model performs well in medical image segmentation. In this paper, UNet model is improved by the same padding after each convolution, so that the image scale remains unchanged through convolution, and the edges of the image are no longer cut off. The improved UNet model is trained for semantic segmentation of the liver in the portal vein in CT images, using binary cross entropy as the loss function, and dice value as the performance evaluation index. The average dice value of the test set reaches 0.85. Our work can be used to help for daily work of liver image segmentation.

Original language	English
Article number	9338815
Pages (from-to)	2315-2318
Number of pages	4
Journal	ITAIC 2020 - IEEE 9th Joint International Information Technology and Artificial Intelligence Conference
DOIs	https://doi.org/10.1109/ITAIC49862.2020.9338815
Publication status	Published - 2020
Externally published	Yes
Event	9th IEEE Joint International Information Technology and Artificial Intelligence Conference, ITAIC 2020 - Chongqing, China Duration: 11 Dec 2020 → 13 Dec 2020

Keywords

CT image
Insert
Liver segmentation
UNet

Access to Document

10.1109/ITAIC49862.2020.9338815

Cite this

@article{12d84b9fd95d463f8b3468e8b904b8d8,

title = "Liver segemtation in CT image with no-edge-cuting UNet",

abstract = "UNet model performs well in medical image segmentation. In this paper, UNet model is improved by the same padding after each convolution, so that the image scale remains unchanged through convolution, and the edges of the image are no longer cut off. The improved UNet model is trained for semantic segmentation of the liver in the portal vein in CT images, using binary cross entropy as the loss function, and dice value as the performance evaluation index. The average dice value of the test set reaches 0.85. Our work can be used to help for daily work of liver image segmentation.",

keywords = "CT image, Insert, Liver segmentation, UNet",

author = "Wenbing Zhao and Kairan Li and Di Zhao and Yunpeng Jiang and Ji Wu and Jinjun Chen and Xianyue Quan and Xinming Li and Feng Xue",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; 9th IEEE Joint International Information Technology and Artificial Intelligence Conference, ITAIC 2020 ; Conference date: 11-12-2020 Through 13-12-2020",

year = "2020",

doi = "10.1109/ITAIC49862.2020.9338815",

language = "English",

pages = "2315--2318",

journal = "ITAIC 2020 - IEEE 9th Joint International Information Technology and Artificial Intelligence Conference",

issn = "2693-2865",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - Liver segemtation in CT image with no-edge-cuting UNet

AU - Zhao, Wenbing

AU - Li, Kairan

AU - Zhao, Di

AU - Jiang, Yunpeng

AU - Wu, Ji

AU - Chen, Jinjun

AU - Quan, Xianyue

AU - Li, Xinming

AU - Xue, Feng

PY - 2020

Y1 - 2020

N2 - UNet model performs well in medical image segmentation. In this paper, UNet model is improved by the same padding after each convolution, so that the image scale remains unchanged through convolution, and the edges of the image are no longer cut off. The improved UNet model is trained for semantic segmentation of the liver in the portal vein in CT images, using binary cross entropy as the loss function, and dice value as the performance evaluation index. The average dice value of the test set reaches 0.85. Our work can be used to help for daily work of liver image segmentation.

AB - UNet model performs well in medical image segmentation. In this paper, UNet model is improved by the same padding after each convolution, so that the image scale remains unchanged through convolution, and the edges of the image are no longer cut off. The improved UNet model is trained for semantic segmentation of the liver in the portal vein in CT images, using binary cross entropy as the loss function, and dice value as the performance evaluation index. The average dice value of the test set reaches 0.85. Our work can be used to help for daily work of liver image segmentation.

KW - CT image

KW - Insert

KW - Liver segmentation

KW - UNet

UR - http://www.scopus.com/inward/record.url?scp=85101469811&partnerID=8YFLogxK

U2 - 10.1109/ITAIC49862.2020.9338815

DO - 10.1109/ITAIC49862.2020.9338815

M3 - Conference article

AN - SCOPUS:85101469811

SN - 2693-2865

SP - 2315

EP - 2318

JO - ITAIC 2020 - IEEE 9th Joint International Information Technology and Artificial Intelligence Conference

JF - ITAIC 2020 - IEEE 9th Joint International Information Technology and Artificial Intelligence Conference

M1 - 9338815

T2 - 9th IEEE Joint International Information Technology and Artificial Intelligence Conference, ITAIC 2020

Y2 - 11 December 2020 through 13 December 2020

ER -

Liver segemtation in CT image with no-edge-cuting UNet

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this