LUneXt: Simple and Efficient U-shaped Network Design for Medical Image Segmentation with Nonlinear Activation

Medical image segmentation has always been a challenging task. This paper proposes a new LUneXt medical image segmentation model based on the characteristics analysis of medical image data sets and testing of different nonlinear activation units. Both normalized activations for the original negative...

Full description

Saved in:
Bibliographic Details
Main Authors: Guanghong Deng, Bing Yu, Wenlong Jing, Yong Li, Xiaodan Zhao
Format: Article
Language:English
Published: World Scientific Publishing 2024-01-01
Series:Computing Open
Subjects:
Online Access:https://www.worldscientific.com/doi/10.1142/S2972370124500077
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1832542498615984128
author Guanghong Deng
Bing Yu
Wenlong Jing
Yong Li
Xiaodan Zhao
author_facet Guanghong Deng
Bing Yu
Wenlong Jing
Yong Li
Xiaodan Zhao
author_sort Guanghong Deng
collection DOAJ
description Medical image segmentation has always been a challenging task. This paper proposes a new LUneXt medical image segmentation model based on the characteristics analysis of medical image data sets and testing of different nonlinear activation units. Both normalized activations for the original negative input image activation have good optimization capabilities for the tokenization module parameters proposed in the original UneXt model. Using different activation coefficients for different foreground and background areas has achieved better results. The experimental results of this paper on the Breast Ultrasound Images (BUSI) data set reached an intersection over union (IoU) value of 62.64%, a Dice value of 76.12%, and a single inference speed of 807.57[Formula: see text]ms. The experimental IoU value of the International Skin Imaging Collaboration (ISIC 2018) data set reached 82.95%, and the Dice value reached 90.50%. The single inference speed reached 842.58[Formula: see text]ms. The LUneXt model is more robust than other models. While improving model performance, it does not introduce higher computational complexity and does not have a major impact on the processing speed of a single image.
format Article
id doaj-art-281e3815c0284aa48be5715f576143ed
institution Kabale University
issn 2972-3701
language English
publishDate 2024-01-01
publisher World Scientific Publishing
record_format Article
series Computing Open
spelling doaj-art-281e3815c0284aa48be5715f576143ed2025-02-04T03:24:11ZengWorld Scientific PublishingComputing Open2972-37012024-01-010210.1142/S2972370124500077LUneXt: Simple and Efficient U-shaped Network Design for Medical Image Segmentation with Nonlinear ActivationGuanghong Deng0Bing Yu1Wenlong Jing2Yong Li3Xiaodan Zhao4GuangDong Engineering Technology Research Center of UAV Remote Sensing Network, Guangzhou iMapCloud Intelligent Technology Co., Ltd., Guangzhou 510095, P. R. ChinaGuangDong Engineering Technology Research Center of UAV Remote Sensing Network, Guangzhou iMapCloud Intelligent Technology Co., Ltd., Guangzhou 510095, P. R. ChinaGuangdong Province Engineering Laboratory for Geographic Spatiotemporal Big Data, Key Laboratory of Guangdong for Utilization of Remote Sensing and Geographical Information System, Guangdong Open Laboratory of Geospatial Information Technology and Application, Guangzhou Institute of Geography, Guangdong Academy of Sciences, Guangzhou 510070, P. R. ChinaGuangdong Province Engineering Laboratory for Geographic Spatiotemporal Big Data, Key Laboratory of Guangdong for Utilization of Remote Sensing and Geographical Information System, Guangdong Open Laboratory of Geospatial Information Technology and Application, Guangzhou Institute of Geography, Guangdong Academy of Sciences, Guangzhou 510070, P. R. ChinaGuangDong Engineering Technology Research Center of UAV Remote Sensing Network, Guangzhou iMapCloud Intelligent Technology Co., Ltd., Guangzhou 510095, P. R. ChinaMedical image segmentation has always been a challenging task. This paper proposes a new LUneXt medical image segmentation model based on the characteristics analysis of medical image data sets and testing of different nonlinear activation units. Both normalized activations for the original negative input image activation have good optimization capabilities for the tokenization module parameters proposed in the original UneXt model. Using different activation coefficients for different foreground and background areas has achieved better results. The experimental results of this paper on the Breast Ultrasound Images (BUSI) data set reached an intersection over union (IoU) value of 62.64%, a Dice value of 76.12%, and a single inference speed of 807.57[Formula: see text]ms. The experimental IoU value of the International Skin Imaging Collaboration (ISIC 2018) data set reached 82.95%, and the Dice value reached 90.50%. The single inference speed reached 842.58[Formula: see text]ms. The LUneXt model is more robust than other models. While improving model performance, it does not introduce higher computational complexity and does not have a major impact on the processing speed of a single image.https://www.worldscientific.com/doi/10.1142/S2972370124500077Image segmentationmedical imagesneural networksactivation function
spellingShingle Guanghong Deng
Bing Yu
Wenlong Jing
Yong Li
Xiaodan Zhao
LUneXt: Simple and Efficient U-shaped Network Design for Medical Image Segmentation with Nonlinear Activation
Computing Open
Image segmentation
medical images
neural networks
activation function
title LUneXt: Simple and Efficient U-shaped Network Design for Medical Image Segmentation with Nonlinear Activation
title_full LUneXt: Simple and Efficient U-shaped Network Design for Medical Image Segmentation with Nonlinear Activation
title_fullStr LUneXt: Simple and Efficient U-shaped Network Design for Medical Image Segmentation with Nonlinear Activation
title_full_unstemmed LUneXt: Simple and Efficient U-shaped Network Design for Medical Image Segmentation with Nonlinear Activation
title_short LUneXt: Simple and Efficient U-shaped Network Design for Medical Image Segmentation with Nonlinear Activation
title_sort lunext simple and efficient u shaped network design for medical image segmentation with nonlinear activation
topic Image segmentation
medical images
neural networks
activation function
url https://www.worldscientific.com/doi/10.1142/S2972370124500077
work_keys_str_mv AT guanghongdeng lunextsimpleandefficientushapednetworkdesignformedicalimagesegmentationwithnonlinearactivation
AT bingyu lunextsimpleandefficientushapednetworkdesignformedicalimagesegmentationwithnonlinearactivation
AT wenlongjing lunextsimpleandefficientushapednetworkdesignformedicalimagesegmentationwithnonlinearactivation
AT yongli lunextsimpleandefficientushapednetworkdesignformedicalimagesegmentationwithnonlinearactivation
AT xiaodanzhao lunextsimpleandefficientushapednetworkdesignformedicalimagesegmentationwithnonlinearactivation