LeaNet: Lightweight U-shaped architecture for high-performance skin cancer image segmentation

计算机科学 分割 图像分割 特征(语言学) 人工智能 像素 模式识别(心理学) 计算机视觉 哲学 语言学
作者
Binbin Hu,Pan Zhou,Hongfang Yu,Yueyue Dai,Ming Wang,Shengbo Tan,Ying Sun
出处
期刊:Computers in Biology and Medicine [Elsevier BV]
卷期号:169: 107919-107919 被引量:29
标识
DOI:10.1016/j.compbiomed.2024.107919
摘要

Skin cancer diagnosis often relies on image segmentation as a crucial aid, and a high-performance segmentation can lower misdiagnosis risks. Part of the medical devices often have limited computing power for deploying image segmentation algorithms. However, existing high-performance algorithms for image segmentation primarily rely on computationally intensive large models, making it challenging to meet the lightweight deployment requirement of medical devices. State-of-the-art lightweight models are not able to capture both local and global feature information of lesion edges due to their model structures, result in pixel loss of lesion edge. To tackle this problem, we propose LeaNet, a novel U-shaped network for high-performance yet lightweight skin cancer image segmentation. Specifically, LeaNet employs multiple attention blocks in a lightweight symmetric U-shaped design. Each blocks contains a dilated efficient channel attention (DECA) module for global and local contour information and an inverted external attention (IEA) module to improve information correlation between data samples. Additionally, LeaNet uses an attention bridge (AB) module to connect the left and right sides of the U-shaped architecture, thereby enhancing the model's multi-level feature extraction capability. We tested our model on ISIC2017 and ISIC2018 datasets. Compared with large models like ResUNet, LeaNet improved the ACC, SEN, and SPEC metrics by 1.09 %, 2.58 %, and 1.6 %, respectively, while reducing the model's parameter number and computational complexity by 570x and 1182x. Compared with lightweight models like MALUNet, LeaNet achieved improvements of 2.07 %, 4.26 %, and 3.11 % in ACC, SEN, and SPEC, respectively, reducing the parameter number and computational complexity by 1.54x and 1.04x.
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
更新
PDF的下载单位、IP信息已删除 (2025-6-4)

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
rioo发布了新的文献求助10
刚刚
天天快乐应助Dowe采纳,获得10
1秒前
YW发布了新的文献求助10
2秒前
上官若男应助给大佬递茶采纳,获得10
2秒前
搜集达人应助WuYujie采纳,获得30
2秒前
3秒前
ZHAO完成签到,获得积分20
4秒前
布鲁爱思发布了新的文献求助10
4秒前
Rondab应助三水采纳,获得10
6秒前
huayi发布了新的文献求助10
6秒前
彭于晏应助zhangtengteng采纳,获得10
7秒前
8秒前
量子星尘发布了新的文献求助10
8秒前
搜集达人应助YW采纳,获得10
8秒前
8秒前
9秒前
布鲁爱思完成签到,获得积分10
10秒前
CipherSage应助yao采纳,获得10
11秒前
情怀应助lx采纳,获得10
11秒前
ZHAO发布了新的文献求助10
12秒前
坚果发布了新的文献求助10
14秒前
biyeshunli发布了新的文献求助10
14秒前
斯文败类应助受伤幻桃采纳,获得10
15秒前
Jasper应助海德堡采纳,获得10
17秒前
17秒前
张先森完成签到,获得积分10
18秒前
19秒前
啊噢完成签到,获得积分10
19秒前
扎心发布了新的文献求助10
20秒前
biyeshunli完成签到,获得积分20
21秒前
Hello应助llm采纳,获得10
22秒前
lx发布了新的文献求助10
22秒前
22秒前
23秒前
23秒前
传奇3应助Mark采纳,获得10
26秒前
28秒前
Andrew发布了新的文献求助10
28秒前
漂亮的秋天完成签到,获得积分10
28秒前
海德堡发布了新的文献求助10
29秒前
高分求助中
A new approach to the extrapolation of accelerated life test data 1000
ACSM’s Guidelines for Exercise Testing and Prescription, 12th edition 500
Picture Books with Same-sex Parented Families: Unintentional Censorship 500
Nucleophilic substitution in azasydnone-modified dinitroanisoles 500
不知道标题是什么 500
A Preliminary Study on Correlation Between Independent Components of Facial Thermal Images and Subjective Assessment of Chronic Stress 500
Phylogenetic study of the order Polydesmida (Myriapoda: Diplopoda) 370
热门求助领域 (近24小时)
化学 材料科学 医学 生物 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 遗传学 基因 物理化学 催化作用 冶金 细胞生物学 免疫学
热门帖子
关注 科研通微信公众号,转发送积分 3971720
求助须知:如何正确求助?哪些是违规求助? 3516377
关于积分的说明 11182327
捐赠科研通 3251591
什么是DOI,文献DOI怎么找? 1795960
邀请新用户注册赠送积分活动 876171
科研通“疑难数据库(出版商)”最低求助积分说明 805340