SegRWKV: Fast and accurate biomedical image segmentation via a receptance-weighted key-value network

计算机科学 特征(语言学) 分割 人工智能 背景(考古学) 架空(工程) 编码器 卷积神经网络 可扩展性 图像分割 计算复杂性理论 推论 计算机视觉 特征提取 模式识别(心理学) 解码方法 依赖关系(UML) 极限(数学) 深度学习 图像(数学) 图像压缩 图像分辨率 面子(社会学概念) 加速 编码(内存) 频道(广播) 医学影像学
作者
Tianyu Cao,Mingyao Ma,Haoxiang Zhang,Guoyuan Zhou,Ruiheng Li
出处
期刊:Journal of Computational Design and Engineering [Oxford University Press]
卷期号:13 (8): 196-211 被引量:1
标识
DOI:10.1093/jcde/qwag065
摘要

Abstract Medical image segmentation is critical for clinical diagnosis, treatment planning, and prognosis. However, practical deployment of segmentation models is often hindered by two main challenges: (1) inefficient global dependency modelling, which slows down context integration and limits feature expressiveness; and (2) insufficient decoder feature reconstruction and multi-scale alignment, which prevents accurate recovery of fine anatomical details. These challenges are exacerbated in high-resolution medical imaging, where computational complexity and slow inference speed further limit model effectiveness. Current solutions, including convolutional neural network (CNN-), transformer-, and state-space-based models, fail to simultaneously balance efficiency, fine-detail reconstruction, and scalability. To address these limitations, we propose SegRWKV(Segmentation Receptance-Weighted Key Value), a hybrid architecture that integrates pretrained vision-RWKV (VRWKV) modules in the encoder for efficient global dependency modelling, and a PixelRefinement module in the decoder to improve feature reconstruction and multi-scale alignment. By using recursive spatial mixing and channel-wise decoupled projections, SegRWKV reduces computational overhead while maintaining high feature expressiveness. A lightweight variant, SegRWKV-S, reduces model size by decreasing channel width and convolutional depth, achieving compression without sacrificing accuracy. Comprehensive experiments on several medical imaging data sets, including AbdomenMRI, endoscopy, and microscopy, demonstrate that SegRWKV consistently outperforms CNN-, vision transformer-, Mamba-, and RWKV-UNet-based baselines. Specifically, SegRWKV-B improves DSC/F1 scores by 4–22% over CNNs, 8–58% over transformers, 1–8% over Mamba, and 10–27% over RWKV-UNet. Efficiency analysis shows that SegRWKV delivers $4.8\,\mathrm{ times}$ faster inference and reduces GPU memory usage by $4.5\,\mathrm{ times}$ compared to existing architectures, with stable computational scaling as input resolution increases. These results highlight SegRWKV’s superior ability to balance accuracy, efficiency, and scalability, particularly in high-resolution medical image segmentation tasks, where many attention-based models struggle to scale effectively. Code and pretrained models are available at: https://github.com/luminescentfrr/SegRWKV/.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
STARRY完成签到,获得积分10
刚刚
1秒前
Rui完成签到,获得积分10
1秒前
烟花的应助被风吹麦浪采纳,获得10
2秒前
orixero的应助被风吹麦浪采纳,获得10
2秒前
2秒前
2秒前
Exile完成签到,获得积分20
2秒前
3秒前
Jessie发布了新的文献求助20
3秒前
haonanchen发布了新的文献求助10
4秒前
kankan84发布了新的文献求助10
4秒前
luoyun完成签到,获得积分10
4秒前
DW的应助被热情的乐荷采纳,获得10
5秒前
bkagyin的应助被小小孟德斯鸠采纳,获得10
5秒前
echooo完成签到,获得积分10
6秒前
6秒前
漂亮的含双完成签到,获得积分10
6秒前
6秒前
快去爬山完成签到 ,获得积分10
6秒前
7秒前
JOJO完成签到,获得积分10
7秒前
better完成签到,获得积分10
7秒前
不安的煜城完成签到,获得积分10
7秒前
8秒前
搜集达人的应助被forever采纳,获得10
8秒前
燕子发布了新的文献求助10
8秒前
zhs完成签到,获得积分10
8秒前
周芷卉完成签到 ,获得积分10
9秒前
9秒前
会飞的猪完成签到,获得积分10
9秒前
9秒前
我要飞完成签到,获得积分10
10秒前
10秒前
Hana完成签到,获得积分20
10秒前
小羊肖恩完成签到,获得积分20
11秒前
11秒前
无限翅膀发布了新的文献求助10
12秒前
12秒前
12秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
拉伸高分子熔体中缠结行为的模拟和小角中子散射研究 1700
奈特基础组织学:配套组织病理学(第4版)(Netter's Essential Histology With Correlated Histopathology) 800
John McDowells Bildungstheorie 800
八旗制度の研究 600
plantTFDB 508
高分子基功能界面水蒸发材料与应用研究 500
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 工程类 有机化学 化学工程 生物化学 物理 复合材料 内科学 催化作用 光电子学 人工智能 细胞生物学 无机化学 物理化学 数学
热门帖子
关注 科研通微信公众号,转发送积分 8176432
求助须知:如何正确求助?哪些是违规求助? 9567768
关于积分的说明 21794704
捐赠科研通 7771112
什么是DOI,文献DOI怎么找? 3431998
关于科研通互助平台的介绍 2556324
邀请新用户注册赠送积分活动 2477671