Multi-Scale Network for Thoracic Organs Segmentation

作     者：Muhammad Ibrahim Khalil Samabia Tehsin Mamoona Humayun N.Z Jhanjhi Mohammed A.AlZain 

作者机构：Department of Computer ScienceBahria UniversityIslamabadPakistan Department of Information systemsCollege of Computer and Information SciencesJouf UniversityKSA School of Computer Science and Engineering(SCE)Taylor’s UniversityMalaysia Center for Smart Society 5.0(CSS5)Faculty of Innovation and Technology(FIT)Taylor’s UniversityMalaysia Department of Information TechnologyCollege of Computers and Information TechnologyTaif UniversityTaif21944Saudi Arabia 

出 版 物：《Computers, Materials & Continua》 (计算机、材料和连续体（英文）)

年 卷 期：2022年第70卷第2期

页      面：3251-3265页

核心收录：

学科分类：0831[工学-生物医学工程（可授工学、理学、医学学位）] 1002[医学-临床医学] 100210[医学-外科学(含：普外、骨外、泌尿外、胸心外、神外、整形、烧伤、野战外)] 0805[工学-材料科学与工程（可授工学、理学学位）] 0801[工学-力学（可授工学、理学学位）] 0812[工学-计算机科学与技术（可授工学、理学学位）] 10[医学] 

基　　金：Taif University Researchers Supporting Project number(TURSP-2020/98) Taif University Taif Saudi Arabia 

主　　题：Deep learning convolutional neural network computed tomography organs at risk computer-aided diagnostic 

摘      要：Medical Imaging Segmentation is an essential technique for modern medical *** is the foundation of many aspects of clinical diagnosis,oncology,and computer-integrated surgical *** significant successes have been achieved in the segmentation of medical images,DL(deep learning)*** delineation of OARs(organs at risk)is vastly dominant but it is prone to errors given the complex irregularities in shape,low texture diversity between tissues and adjacent blood area,patientwide location of organisms,and weak soft tissue contrast across adjacent organs in CT *** now several models have been implemented onmulti organs segmentation but not caters to the problemof imbalanced classes some organs have relatively small pixels as compared to *** segment OARs in thoracic CT images,we proposed the model based on the encoder-decoder approach using transfer learning with the efficientnetB7 DL *** have built a fully connected CNN(Convolutional Neural network)having 5 layers of encoding and 5 layers of decoding with efficientnetB7 specifically to tackle imbalance class pixels in an accurate way for the segmentation of *** methodology achieves 0.93405 IOU score,0.95138 F1 score and class-wise dice score for esophagus 0.92466,trachea 0.94257,heart 0.95038,aorta 0.9351 and background *** results showed that our proposed framework can be segmented organs accurately.