改进U-Net的X射线乳腺肿瘤区域分割方法

打开文本图片集

X-ray Breast Tumor Region Segmentation Method Using an Improved U-Net

CHEN Xiang1CAI Yanguang1，2 GONG Guojun³ZHANG Ruihu4 CAI Hao5

（1.College of Automation， Guangdong University of Technology， Guangzhou 51ooo6， China 2.School of Intelligent Manufacturing and Electrical Engineering， Guangzhou Institute of Science and Technology， Guangzhou 510540， China3.Guangdong Lingnan First Technical College of Industry and Commerce， Guangzhou 510000， China4.Guangdong Yuedong Technician Colege， Shantou 515000， China 5.Dongguan Industrial Investment Holding Group Co.， Ltd.， Dongguan 5230oo， China）

Abstract：Breastcancerisaserious disease that threatens women's health，and timelyandaccurate diagnosis iscrucial for reducing its mortalityrate.Toimprovetheacuracyofbreastcancerdiagnosis，thisstudyproposesanenhanced U-Net-based method forsegmentingbasttmregiosinX-ryimags.Fist，pralelaetiomouleisostructedbyitegatingsatialdael atentionmechanisms tostrengthenthenetwork'sabilitytoextractandidentifykeyfeatures.Next，theparalelatentionoduleis combinedwitharesidalmoduletodesignaresidualparalleltentionmodule，ehancingtheU-Netmodel'sdeepfeatureetraction andhigh-ffcousgaabilitFallealpalltouleooatdtoderatteUNetmodel，improving itssegmentationacuracyforX-raybreasttumorregions.ExperimentalresultsontheCBIS-DDMdataset demonstrate that the improved U-Net model achieves Dice coeffcients and mean intersection over union （mIoU）of 94.20% and 90.76% ，respectively， significantly enhancing the segmentation accuracy of X-ray breast tumor regions.

Keywords： image segmentation; attention mechanism; breast tumor; residual network; U-Net

0 引言

乳腺癌是女性常见的恶性肿瘤之一，约占所有女性癌症病例的 30%[1-2] ，占癌症相关死亡总数的 17%[3] 已成为女性癌症死亡的主要原因之一。（剩余9862字）