融合级联Transformer和YOLOv8的膝关节多类别囊肿检测方法研究  

Research on knee multi-class cyst detection algorithm based on cascaded Transformer and YOLOv8

作　　者：张丽媛[1,2] 张驰 蒋振刚 唐雄风[3] ZHANG Liyuan;ZHANG Chi;JIANG Zhengang;TANG Xiongfeng(School of Computer Science and Technology,Changchun University of Science and Technology,Changchun 130022,China;Digital Healthcare Research Center,Zhongshan Institute of Changchun University of Science and Technology,Zhongshan 528436,China;Orthopedic Medical Center,Jilin University Second Hospital,Changchun 130041)

机构地区：[1]长春理工大学计算机科学技术学院,长春130022 [2]长春理工大学中山研究院数字医疗研究中心,中山528436 [3]吉林大学第二医院骨科医疗中心,长春130041

出　　处：《生物医学工程研究》2025年第1期58-66,共9页Journal Of Biomedical Engineering Research

基　　金：国家自然科学基金项目(U21A20390);吉林省教育厅项目(JJKH20240945KJ)。

摘　　要：针对膝关节囊肿磁共振(MR)影像中囊肿与关节内积液及其他组织特征相似性高,边界模糊的问题,本研究提出了一种膝关节囊肿病变检测模型YOLO-Cyst。首先,在骨干网络部分采用级联Vision Transformer模块获取长距离上下文信息,提高囊肿检测的准确性;其次,在YOLOv8的跨阶段部分连接与双融合模块中引入可变形大核注意力模块,增强模型的局部特征提取能力。实验结果表明:与YOLOv8相比,YOLO-Cyst的mAP50和mAP50-95指标分别提高了5.1%和0.8%;与Faster R-CNN和DETR相比,YOLO-Cyst的mAP50指标分别提高了23.9%和13.0%,mAP50-95指标分别提升了10.7%和6.8%。本研究所提算法能有效学习丰富的膝关节囊肿特征表示,实现对不同类型和形态的囊肿的精确检测。Aiming at the high similarity and blurred boundary between the cysts and intra-articular fluid and other tissues in magnetic resonance(MR)images of knee cysts,we proposed a knee cysts lesion detection model YOLO-Cyst.Firstly,in the backbone network,a cascaded Vision Transformer module was employed to capture long-distance contextual information,thereby enhancing cyst detection accuracy.Secondly,building upon the cross-stage partial connectivity and dual fusion module of YOLOv8,a deformable large kernel attention module was introduced to enhance the model capacity for local feature extraction.Experimental results demonstrated that compared to the YOLOv8,YOLO-Cyst improved mAP50 and mAP50-95 by 5.1%,0.8%,respectively.Furthermore,compared to the Faster R-CNN and DETR,YOLO-Cyst enhanced mAP50 by 23.9%,13.0%,and mAP50-95 by 10.7%,6.8%,respectively.This algorithm can learn rich feature representation of knee cysts and enable accurate detection of cysts of different types and morphologies.

关 键 词：膝关节囊肿 目标检测 上下文信息 TRANSFORMER YOLOv8 

分 类 号：R318[医药卫生—生物医学工程] TP18[医药卫生—基础医学]