DMR-KAN：基于多尺度区域强化的三维肿瘤影像分割方法

方丁毅; 程换新; 骆晓玲; 王若峥

doi:10.12263/DZXB.20250379

您当前的位置：

首页 >

文章列表页 >

DMR-KAN：基于多尺度区域强化的三维肿瘤影像分割方法

学术论文 | 更新时间：2025-12-27

- DMR-KAN：基于多尺度区域强化的三维肿瘤影像分割方法
- DMR-KAN: A 3D Medical Image Segmentation Method Based on Multi-Scale Region Enhancement
- 电子学报 2025年53卷第8期页码：2818-2829
- 作者机构：
  
  1.青岛科技大学自动化与电子工程学院，山东青岛 266061
  2.青岛科技大学机电工程学院，山东青岛 266061
  3.新疆医科大学附属肿瘤医院放疗中心，新疆乌鲁木齐 830022
- 作者简介：
  
  [ "方丁毅男，1999年6月生，黑龙江嫩江人.青岛科技大学自动化与电子工程学院硕士研究生.主要研究方向为人工智能与机器视觉.E-mail: 2682758035@qq.com" ]
  [ "程换新男，1966年9月生，安徽安庆人.青岛科技大学自动化与电子工程学院教授，硕士生导师.主要研究方向为机器视觉、智能控制系统、深度学习理论及应用等. E-mail: 2718228346@qq.com" ]
  [ "骆晓玲女，1966年5月生，山东菏泽人.青岛科技大学机电工程学院教授、硕士生导师.主要研究方向为过程装备及自动化的优化设计、深度学习参数优化与过程控制. E-mail: luoxiaoling@qust.edu.cn" ]
  [ "王若峥女，1964年1月生，辽宁营口人.博士后、教授、主任医师（二级）、博士生导师，国务院政府特殊津贴专家.主要研究方向为肿瘤放疗基础及临床.E-mail: wrz8526@vip.163.com" ]
- 基金信息：
- DOI：10.12263/DZXB.20250379
  中图分类号： TP391.41;
- 收稿：2025-05-14，
  
  录用：2025-08-04，
  
  纸质出版：2025-08-25
- 稿件说明：
移动端阅览
方丁毅, 程换新, 骆晓玲, 等. DMR-KAN：基于多尺度区域强化的三维肿瘤影像分割方法[J]. 电子学报, 2025, 53(08): 2818-2829.

FANG Ding-yi, CHENG Huan-xin, LUO Xiao-ling, et al. DMR-KAN: A 3D Medical Image Segmentation Method Based on Multi-Scale Region Enhancement[J]. Acta Electronica Sinica, 2025, 53(08): 2818-2829.
方丁毅, 程换新, 骆晓玲, 等. DMR-KAN：基于多尺度区域强化的三维肿瘤影像分割方法[J]. 电子学报, 2025, 53(08): 2818-2829. DOI：10.12263/DZXB.20250379

FANG Ding-yi, CHENG Huan-xin, LUO Xiao-ling, et al. DMR-KAN: A 3D Medical Image Segmentation Method Based on Multi-Scale Region Enhancement[J]. Acta Electronica Sinica, 2025, 53(08): 2818-2829. DOI：10.12263/DZXB.20250379

摘要

KAN（Kolmogorov-Arnold Networks）模型通过新的线性函数拟合方式使得图像分割的精准度得到提升.然而，其拟合角度单一、提取标签位置信息差等问题，导致其对标签细节特征信息的处理能力较差，令网络精度提升受限.针对上述问题，设计了多尺度双通道三维影像分割模型，该模型通过整合多角度三维影像输入，将多角度KAN模块与多尺度卷积加权残差通道相结合，显著增强了网络对图像微小特征的提取能力.在网络注意力机制方面，设计了多视角自注意力残差模块，该模块通过多维度特征交互有效捕获标签空间位置信息，使占比较低（<10%）的标签区域仍能保持优异的分割精度.模型在BraTS2021 MRI多模态三维脑肿瘤数据集与LiTS2017肝肿瘤CT三维数据集中进行实验.改进模型精准度分别为86.54%与88.07%；在脑肿瘤数据集中，增强肿瘤、全部肿瘤和肿瘤核心区域的Dice评价指标达到83.67%、88.79%和85.28%，相较U-KAN网络分别提升3.38、2.85和1.62个百分点；在肝肿瘤数据集中，肝脏与肿瘤区域的Dice评价指标达到91.36%与84.77%，分别提升了1.69个百分点与1.02个百分点.实验结果表明该模型对三维肿瘤影像分割效果提升显著.

Abstract

The KAN (Kolmogorov-Arnold Networks) model enables the accuracy of image segmentation to be improved by a new linear function fitting method. However

the problems of single fitting angle and poor extraction of label position information lead to its poor ability to process the detailed feature information of labels

which limits the improvement of network accuracy. To address the above problems

a multi-scale dual-channel 3D image segmentation model is designed

which significantly enhances the network’s ability to extract minute features from images by integrating multi-angle 3D image inputs and combining the multi-angle KAN module with multi-scale convolutional weighted residual channels. In terms of the network attention mechanism

a multi-view self-attentive residual module is designed

which effectively captures the label spatial location information through multi-dimensional feature interactions

so that the label region with a relatively low percentage (<10%) can still maintain excellent segmentation accuracy. The model is experimented on BraTS2021 MRI multimodal 3D brain tumor dataset and LiTS2017 liver tumor CT 3D dataset. The accuracy of the improved model is 86.54% and 88.07%

respectively; in the brain tumor dataset

the Dice evaluation indexes of the enhanced tumor

all tumors

and tumor core region reach 83.67%

88.79%

and 85.28%

which are improved by 3.38

2.85

and 1.62 percentage points

respectively

compared with the U-KAN network; in the liver tumor dataset

the liver and tumor region’s Dice evaluation index reached 91.36% and 84.77%

which were improved by 1.69 percentage points and 1.02 percentage points

respectively. The experimental results show that the model improves the effect of 3D tumor image segmentation significantly.

关键词

Keywords

references

郑光远 , 刘峡壁 , 韩光辉 . 医学影像计算机辅助检测与诊断系统综述 [J ] . 软件学报 , 2018 , 29 ( 5 ): 1471 - 1514 .

ZHENG G Y , LIU X B , HAN G H . Survey on medical image computer aided detection and diagnosis systems [J ] . Journal of Software , 2018 , 29 ( 5 ): 1471 - 1514 . (in Chinese)

陈正阳 , 陈平华 , 陈建平 . 一种改进的Unet医学影像分割模型 [J ] . 计算机仿真 , 2024 , 41 ( 6 ): 216 - 222 .

CHEN Z Y , CHEN P H , CHEN J P . An improved unet medical image segmentation model [J ] . Computer Simulation , 2024 , 41 ( 6 ): 216 - 222 . (in Chinese)

刘金平 , 吴娟娟 , 张荣 , 等 . 基于结构重参数化与多尺度深度监督的COVID-19胸部CT图像自动分割 [J ] . 电子学报 , 2023 , 51 ( 5 ): 1163 - 1171 .

LIU J P , WU J J , ZHANG R , et al . Toward automated segmentation of COVID-19 chest CT images based on structural reparameterization and multi-scale deep supervision [J ] . Acta Electronica Sinica , 2023 , 51 ( 5 ): 1163 - 1171 . (in Chinese)

CUI M T , LI K , CHEN J Y , et al . CM-unet: A novel remote sensing image segmentation method based on improved U-net [C ] // IEEE Access . Piscataway : IEEE , 2023 : 56994 - 57005 .

KRIZHEVSKY A , SUTSKEVER I , HINTON G E . ImageNet classification with deep convolutional neural networks [J ] . Communications of the ACM , 2017 , 60 ( 6 ): 84 - 90 .

WANG M H , YUAN P , QIU S , et al . Dual-encoder UNet-based narrowband uncooled infrared imaging denoising network [J ] . Sensors , 2025 , 25 ( 5 ): 1476 .

GUO M H , XU T X , LIU J J , et al . Attention mechanisms in computer vision: A survey [J ] . Computational Visual Media , 2022 , 8 ( 3 ): 331 - 368 .

TAY Y , DEHGHANI M , BAHRI D , et al . Efficient transformers: A survey [J ] . ACM Computing Surveys , 2022 , 55 ( 6 ): 1 - 28 .

ZHANG J X , JIANG Z K , DONG J , et al . Attention gate ResU-net for automatic MRI brain tumor segmentation [J ] . IEEE Access , 2020 , 8 : 58533 - 58545 .

VARSHANEYA V , BALASUBRAMANIAN S , DARSHAN G . Res‐SE‐Net: Boosting performance of ResNets by enhancing bridge connections [M ] // Machine Learning Algorithms and Applications , Hoboken : Wiley , 2021 : 61 - 75 .

DAO T , FU D Y , ERMON S , et al . FlashAttention: Fast and memory-efficient exact attention with IO-awareness [EB/OL ] . ( 2022-06-23 )[ 2025-05-10 ] . https://arXiv.org/abs/2205.14135 https://arXiv.org/abs/2205.14135 .

LIU Z M , WANG Y X , VAIDYA S , et al . KAN: Kolmogorov-Arnold networks [EB/OL ] . ( 2025-02-09 )[ 2025-05-10 ] . https://arXiv.org/abs/2404.19756 https://arXiv.org/abs/2404.19756 .

MORADZADEH A , WAWRZYNIAK L , MACKLIN M , et al . UKAN: Unbound Kolmogorov-Arnold network accompanied with accelerated library [EB/OL ] . ( 2024-10-01 )[ 2025-05-10 ] . https://arXiv.org/abs/2408.11200 https://arXiv.org/abs/2408.11200 .

LOU A G , GUAN S Y , LOEW M H . DC-UNet: Rethinking the U-Net architecture with dual channel efficient CNN for medical image segmentation [C ] // Medical Imaging 2021: Image Processing . California : SPIE , 2021 : 98 .

LI C X , LIU X Y , LI W Y , et al . U-KAN makes strong backbone for medical image segmentation and generation [J ] . Proceedings of the AAAI Conference on Artificial Intelligence , 2025 , 39 ( 5 ): 4652 - 4660 .

MENG W J , LIU S J , WANG H J . AFC-Unet: Attention-fused full-scale CNN-transformer unet for medical image segmentation [J ] . Biomedical Signal Processing and Control , 2025 , 99 : 106839 - 106849 .

FANG C L , WU K G . KANDU-Net: A dual-channel U-Net with KAN for medical image segmentation [EB/OL ] . ( 2024-09-30 )[ 2025-05-10 ] . https://arXiv.org/abs/2409.20414 https://arXiv.org/abs/2409.20414 .

AZAD R , NIGGEMEIER L , HÜTTEMANN M , et al . Beyond self-attention: Deformable large kernel attention for medical image segmentation [C ] // 2024 IEEE/CVF Winter Conference on Applications of Computer Vision . Piscataway : IEEE , 2024 : 1276 - 1286 .

WANG C , WANG L , WANG N Q , et al . CFATransUnet: Channel-wise cross fusion attention and transformer for 2D medical image segmentation [J ] . Computers in Biology and Medicine , 2024 , 168 : 107803 .

REN J T , HOCHREUTER K , KALLEHAUGE J F , et al . UMamba adjustment: Advancing GTV segmentation for head and neck cancer in MRI-guided RT with UMamba and NnU-net ResEnc planner [C ] // Head and Neck Tumor Segmentation for MR-Guided Applications . New York : ACM , 2025 : 123 - 135 .

LIU J R , YANG H , ZHOU H Y , et al . Swin-UMamba: Mamba-based UNet with ImageNet-based pretraining [C ] // Medical Image Computing and Computer Assisted Intervention-MICCAI 2024 . New York : ACM , 2024 : 615 - 625 .

ATEK S , MEHIDI I , JABRI D , et al . SwinT-Unet: Hybrid architecture for medical image segmentation based on swin transformer block and dual-scale information [C ] // 2022 7th International Conference on Image and Signal Processing and Their Applications . Piscataway : IEEE , 2022 : 1 - 6 .

LIU A L , JIA D Q , SUN K C , et al . LM-UNet: Whole-body pet-CT lesion segmentation with dual-modality-based annotations driven by Latent mamba U-Net [C ] // Medical Image Computing and Computer Assisted Intervention-MICCAI 2024 . Cham : Springer , 2024 : 405 - 414 .

ZHANG R H , ZOU R Z , ZHAO Y , et al . BA-Net: Bridge attention in deep neural networks [J ] . Expert Systems with Applications , 2024 , 292 : 128525 - 128538 .

CHEN J D , SHEN X Y , ZHAO Y , et al . Attention gate and dilation U-shaped network (GDUNet): An efficient breast ultrasound image segmentation network with multiscale information extraction [J ] . Quantitative Imaging in Medicine and Surgery , 2024 , 14 ( 2 ): 2034 - 2048 .

严武军 , 李建昌 , 叶金霞 . 轻量化GEC-3DUnet骶髂关节影像自动分割 [J ] . 计算机技术与发展 , 2025 , 35 ( 2 ): 33 - 40 .

YAN W J , LI J C , YE J X . Lightweight GEC-3DUnet based automatic segmentation of sacroiliac joint medical imaging [J ] . Computer Technology and Development , 2025 , 35 ( 2 ): 33 - 40 . (in Chinese)

ZHONG H , WU C . T-UNet: Triplet UNet for change detection in high-resolution remote sensing images [J ] . Geo-spatial Information Science , 2025 , 28 ( 2 ): 437 - 454 .

KIM D , YOO Y , PARK S , et al . SelfReg: Self-supervised contrastive regularization for domain generalization [C ] // 2021 IEEE/CVF International Conference on Computer Vision . Piscataway : IEEE , 2022 : 9599 - 9608 .

ISENSEE F , JAEGER P F , KOHL S A A , et al . nnU-Net: A self-configuring method for deep learning-based biomedical image segmentation [J ] . Nature Methods , 2021 , 18 ( 2 ): 203 - 211 .

TONG X Z , WEI J Y , SUN B , et al . ASCU-Net: Attention gate, spatial and channel attention U-Net for skin lesion segmentation [J ] . Diagnostics , 2021 , 11 ( 3 ): 501 .

CHEN J N , MEI J R , LI X H , et al . TransUNet: Rethinking the U-Net architecture design for medical image segmentation through the lens of transformers [J ] . Medical Image Analysis , 2024 , 97 : 103280 .

KARIMI A , FAEZ K , NAZARI S . DEU-Net: Dual-encoder U-Net for automated skin lesion segmentation [J ] . IEEE Access , 2023 , 11 : 134804 - 134821 .

CAO H , WANG Y Y , CHEN J , et al . Swin-Unet: Unet-like pure Transformer for medical image segmentation [M ] // Computer Vision-ECCV 2022 Workshops . Cham : Springer Nature Switzerland , 2023 : 205 - 218 .

XING Z H , YE T , YANG Y J , et al . SegMamba: Long-range sequential modeling mamba for 3D medical image segmentation [C ] // Medical Image Computing and Computer Assisted Intervention-MICCAI 2024 . Cham : Springer , 2024 : 578 - 588 .

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于时频注意力Conformer的多尺度短语音说话人识别模型

ConvFormer：基于Transformer的视觉主干网络

图像压缩感知的特征域优化及自注意力增强神经网络重构算法

SAU-Net：基于U-Net和自注意力机制的医学图像分割方法