计算机与现代化Issue(3)：72-77,6.DOI:10.3969/j.issn.1006-2475.2024.03.012
融合多尺度空间特征的甲状腺结节超声图像分割
Ultrasound Image Segmentation of Thyroid Nodules by Fusing Multi-scale Spatial Features
摘要
Abstract
The ultrasound images of thyroid nodules have serious noise and low contrast between different tissues.The existing ul-trasound image segmentation algorithm of thyroid nodules have some problems of blurred edge information and inaccurate segmen-tation of small nodules.Therefore this paper proposes an ultrasound image segmentation algorithm of thyroid nodules fused with multi-scale spatial features.Based on the U-Net model,the coordinate attention mechanism is introduced to embed the position information into the channel attention to achieve the model's localization of the thyroid nodule region in the coding part.At the same time,the fused multiscale feature module extracts the spatial aspect features.To retain more detailed features,we uses con-volution operation in the process of down sampling and the binary cross-entropy loss and Dice coefficient loss as the comprehen-sive loss.The experimental results show that compared with the benchmark model U-Net,the proposed algorithm model im-proves the F1 evaluation index by 9.9 percentage points,and the accuracy rate is increased to 92.8%.Thus the feasibility and ef-fectiveness is verified.关键词
甲状腺结节/U-Net/空洞卷积/多尺度特征/坐标注意力Key words
thyroid nodule/U-Net/atrous convolution/multi-scale features/coordinate attention分类
信息技术与安全科学引用本文复制引用
崔少国,张宇楠..融合多尺度空间特征的甲状腺结节超声图像分割[J].计算机与现代化,2024,(3):72-77,6.基金项目
国家自然科学基金资助项目(62003065) （62003065）
重庆市科技局自然基金资助项目(2022NSCQ-MSX2933,2022TFII-OFX0262,cstc2019jscx-mbdxX0061) （2022NSCQ-MSX2933,2022TFII-OFX0262,cstc2019jscx-mbdxX0061）
教育部人文社科规划基金资助项目(22YJA870005) （22YJA870005）
重庆市教委重点项目(KJZD-K202200510) （KJZD-K202200510）
重庆市社会科学规划项目(2022NDYB119) （2022NDYB119）
重庆师范大学人才基金资助项目(20XLB004) （20XLB004）
