SHPNeXt: A Novel Method of Multi-Scale and Variable Resolution AI-Based Tongue Image Segmentation

In the domain of Traditional Chinese Medicine, accurately segmenting tongue images is fundamental for computer-assisted diagnosis. Yet, current models often falter with images of diverse scales and clarity, impeding their widespread application. To address this challenge, we propose SHPNeXt, an inno...

Full description

Saved in:
Bibliographic Details
Main Authors: Chong-Xiao Peng, Zhi-Jun Gao, Jin-Huan Wang, Xin Yue, Yi Li, Li-Li Sun, Yin-Huan Sun, Fu-Quan Du
Format: Article
Language:English
Published: IEEE 2025-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10938541/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850263192414978048
author Chong-Xiao Peng
Zhi-Jun Gao
Jin-Huan Wang
Xin Yue
Yi Li
Li-Li Sun
Yin-Huan Sun
Fu-Quan Du
author_facet Chong-Xiao Peng
Zhi-Jun Gao
Jin-Huan Wang
Xin Yue
Yi Li
Li-Li Sun
Yin-Huan Sun
Fu-Quan Du
author_sort Chong-Xiao Peng
collection DOAJ
description In the domain of Traditional Chinese Medicine, accurately segmenting tongue images is fundamental for computer-assisted diagnosis. Yet, current models often falter with images of diverse scales and clarity, impeding their widespread application. To address this challenge, we propose SHPNeXt, an innovative network designed to accurately segment tongue images across different scales and resolutions. This model blends PoolFormer and Hire-MLP to adeptly discern both overarching and nuanced details, ensuring accurate segmentation across varying tongue image sizes. Furthermore, SHPNeXt&#x2019;s precision was further enhanced by integrating a Nuclear-Norm Non-negative Matrix Factorization (NMF) approach, which robustly counters noise in lower quality images. Experiments on three benchmark datasets demonstrate SHPNeXt&#x2019;s superior performance, achieving mean Intersection over Union (mIoU) scores of 99.64%, 97.05%, and 96.82%. Balancing efficiency and accuracy, SHPNeXt&#x2019;s architecture comprises 5.984 million parameters and operates at 1.22 GFLOPs, rendering it an exceptionally effective tool for real-world tongue diagnosis in TCM. The code has been released on github: (<uri>https://github.com/Kuanzhaipcx/SHPNeXt.git</uri>).
format Article
id doaj-art-eccb8ad5ded447f199d47d3975d65090
institution OA Journals
issn 2169-3536
language English
publishDate 2025-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-eccb8ad5ded447f199d47d3975d650902025-08-20T01:55:02ZengIEEEIEEE Access2169-35362025-01-0113545045451610.1109/ACCESS.2025.355448710938541SHPNeXt: A Novel Method of Multi-Scale and Variable Resolution AI-Based Tongue Image SegmentationChong-Xiao Peng0https://orcid.org/0009-0009-6636-9779Zhi-Jun Gao1https://orcid.org/0000-0002-9546-0047Jin-Huan Wang2Xin Yue3Yi Li4Li-Li Sun5Yin-Huan Sun6Fu-Quan Du7School of Computer Information and Engineering, Heilongjiang University of Science and Technology, Harbin, ChinaSchool of Computer Information and Engineering, Heilongjiang University of Science and Technology, Harbin, ChinaFirst Affiliated Hospital, Heilongjiang University of Chinese Medicine, Harbin, ChinaSchool of Computer Information and Engineering, Heilongjiang University of Science and Technology, Harbin, ChinaSchool of Computer Information and Engineering, Heilongjiang University of Science and Technology, Harbin, ChinaSchool of Computer Information and Engineering, Heilongjiang University of Science and Technology, Harbin, ChinaSchool of Computer Information and Engineering, Heilongjiang University of Science and Technology, Harbin, ChinaSchool of Computer Information and Engineering, Heilongjiang University of Science and Technology, Harbin, ChinaIn the domain of Traditional Chinese Medicine, accurately segmenting tongue images is fundamental for computer-assisted diagnosis. Yet, current models often falter with images of diverse scales and clarity, impeding their widespread application. To address this challenge, we propose SHPNeXt, an innovative network designed to accurately segment tongue images across different scales and resolutions. This model blends PoolFormer and Hire-MLP to adeptly discern both overarching and nuanced details, ensuring accurate segmentation across varying tongue image sizes. Furthermore, SHPNeXt&#x2019;s precision was further enhanced by integrating a Nuclear-Norm Non-negative Matrix Factorization (NMF) approach, which robustly counters noise in lower quality images. Experiments on three benchmark datasets demonstrate SHPNeXt&#x2019;s superior performance, achieving mean Intersection over Union (mIoU) scores of 99.64%, 97.05%, and 96.82%. Balancing efficiency and accuracy, SHPNeXt&#x2019;s architecture comprises 5.984 million parameters and operates at 1.22 GFLOPs, rendering it an exceptionally effective tool for real-world tongue diagnosis in TCM. The code has been released on github: (<uri>https://github.com/Kuanzhaipcx/SHPNeXt.git</uri>).https://ieeexplore.ieee.org/document/10938541/Tongue image segmentationSHPNeXtmulti-scale variable resolution imagesdeep learning
spellingShingle Chong-Xiao Peng
Zhi-Jun Gao
Jin-Huan Wang
Xin Yue
Yi Li
Li-Li Sun
Yin-Huan Sun
Fu-Quan Du
SHPNeXt: A Novel Method of Multi-Scale and Variable Resolution AI-Based Tongue Image Segmentation
IEEE Access
Tongue image segmentation
SHPNeXt
multi-scale variable resolution images
deep learning
title SHPNeXt: A Novel Method of Multi-Scale and Variable Resolution AI-Based Tongue Image Segmentation
title_full SHPNeXt: A Novel Method of Multi-Scale and Variable Resolution AI-Based Tongue Image Segmentation
title_fullStr SHPNeXt: A Novel Method of Multi-Scale and Variable Resolution AI-Based Tongue Image Segmentation
title_full_unstemmed SHPNeXt: A Novel Method of Multi-Scale and Variable Resolution AI-Based Tongue Image Segmentation
title_short SHPNeXt: A Novel Method of Multi-Scale and Variable Resolution AI-Based Tongue Image Segmentation
title_sort shpnext a novel method of multi scale and variable resolution ai based tongue image segmentation
topic Tongue image segmentation
SHPNeXt
multi-scale variable resolution images
deep learning
url https://ieeexplore.ieee.org/document/10938541/
work_keys_str_mv AT chongxiaopeng shpnextanovelmethodofmultiscaleandvariableresolutionaibasedtongueimagesegmentation
AT zhijungao shpnextanovelmethodofmultiscaleandvariableresolutionaibasedtongueimagesegmentation
AT jinhuanwang shpnextanovelmethodofmultiscaleandvariableresolutionaibasedtongueimagesegmentation
AT xinyue shpnextanovelmethodofmultiscaleandvariableresolutionaibasedtongueimagesegmentation
AT yili shpnextanovelmethodofmultiscaleandvariableresolutionaibasedtongueimagesegmentation
AT lilisun shpnextanovelmethodofmultiscaleandvariableresolutionaibasedtongueimagesegmentation
AT yinhuansun shpnextanovelmethodofmultiscaleandvariableresolutionaibasedtongueimagesegmentation
AT fuquandu shpnextanovelmethodofmultiscaleandvariableresolutionaibasedtongueimagesegmentation