首页|基于三维容积乳腺超声图像的病灶智能筛查方法

基于三维容积乳腺超声图像的病灶智能筛查方法

扫码查看
乳腺三维容积超声作为乳腺疾病筛查的重要手段,能够提供包含冠状切面在内更多的信息,在乳腺疾病诊断中具有重要的应用价值.但其数据量大且易因操作不当引起伪像,对医生造成较大的阅片负担,影响诊断的正确性.因此,提出一种包含乳腺图像质量控制模块的病灶智能筛查方法,对图像采集过程中易出现且易对结果产生影响的多重反射伪像、耦合不良伪像和较大的空载区域进行检测,对合格的数据采用横断面结果为主、矢状面结果为辅的策略检测超声图像中的疑似病灶,辅助医生完成诊断.经测试验证,该方法质量控制模块的准确率为100%,病灶智能筛查模块的病灶检出率为95.5%,能够达到辅助医生进行阅片从而完成病灶筛查和诊断的目的.
An Intelligent Screening Method for Breast Lesions Based on Three-Dimensional Volume Ultrasound Images
As an important means of breast disease screening,breast three-dimensional volume ultrasound can provide more information including coronal section,and has important application value in the diagnosis of breast diseases.However,the amount of data is large and it is easy to cause artifacts due to improper operation,which causes a large burden on doctors and affects the accuracy of diagnosis.Therefore,this paper proposed an intelligent lesion screening method including breast image quality control module to detect the multiple reflection artifacts and poor coupling artifacts and large no-load regions that are easy to appear in the process of image acquisition and affect the results.For qualified data,the cross-sectional results were used as the main strategy,and the sagittal results were used as the auxiliary strategy to detect suspected lesions in ultrasound images to assist doctors to complete the diagnosis.After testing and verification,the accuracy of the quality control module was 100%,and the lesion detection rate of the intelligent lesion screening module was 95.5%,respectively,which could achieve the purpose of assisting doctors to read images to complete lesion screening and diagnosis.

breast volume ultrasoundlesion screeningartifactYOLO

陈升恺、何桂端

展开 >

汕头市中心医院 (广东 汕头 515041)

乳腺容积超声 病灶筛查 伪像 YOLO

2024

中国医疗器械信息
中国医疗器械行业协会

中国医疗器械信息

影响因子:0.375
ISSN:1006-6586
年,卷(期):2024.30(17)