基于密集交互式融合Mask RCNN的肺部PET/CT图像实例分割
Pulmonary PET/CT image instance segmentation based on dense interactive feature fusion Mask RCNN
周涛 1赵雅楠 1陆惠玲 2王亚星 1支力佳3
作者信息
- 1. 北方民族大学 计算机科学与工程学院(银川 750021);北方民族大学 图像图形智能处理国家民委重点实验室(银川 750021)
- 2. 宁夏医科大学 医学信息工程学院(银川 750004)
- 3. 北方民族大学 计算机科学与工程学院(银川 750021)
- 折叠
摘要
正电子发射断层显像/X线计算机体层成像(PET/CT)肺部图像中存在病灶区域特征像素信息少、形状复杂多样,病变与周围组织界限模糊等问题,导致模型对肿瘤病变特征提取不充分.针对上述问题,本文提出基于密集交互式融合Mask RCNN(DIF-Mask RCNN)实例分割模型.首先设计具有跨尺度主辅结构的特征提取网络,提取出不同尺度病灶特征;然后设计密集交互式增强辅助网络(DIFEN),通过将最浅层病变特征以密集连接形式与邻近特征、当前特征进行交互融合,增强深层特征图中病灶细节信息;最后构建密集交互式融合金字塔网络(DIF-FPN),在自下而上路径中将浅层信息逐个以密集连接方式补充到深层特征中,进一步加强模型对病变区域的微弱特征感知力.在临床PET/CT肺部图像数据集上进行消融实验和对比实验,结果表明所提模型对于病变区域实例分割的 APdet、APseg、APdet_s、APseg_s 指标分别为 67.16％、68.12％、34.97％、37.68％,与 Mask RCNN(ResNet50)相比在APdet和APseg指标上分别提升7.11％、5.14％.DIF-Mask RCNN模型能够有效检测分割肿瘤病变,为肺癌辅助诊断提供重要的参考价值与评估依据.
Abstract
There are some problems in positron emission tomography/computed tomography(PET/CT)lung images,such as little information of feature pixels in lesion regions,complex and diverse shapes,and blurred boundaries between lesions and surrounding tissues,which lead to inadequate extraction of tumor lesion features by the model.To solve the above problems,this paper proposes a dense interactive feature fusion Mask RCNN(DIF-Mask RCNN)model.Firstly,a feature extraction network with cross-scale backbone and auxiliary structures was designed to extract the features of lesions at different scales.Then,a dense interactive feature enhancement network was designed to enhance the lesion detail information in the deep feature map by interactively fusing the shallowest lesion features with neighboring features and current features in the form of dense connections.Finally,a dense interactive feature fusion feature pyramid network(FPN)network was constructed,and the shallow information was added to the deep features one by one in the bottom-up path with dense connections to further enhance the model's perception of weak features in the lesion region.The ablation and comparison experiments were conducted on the clinical PET/CT lung image dataset.The results showed that the APdet,APseg,APdet_s and APseg_s indexes of the proposed model were 67.16％,68.12％,34.97％and 37.68％,respectively.Compared with Mask RCNN(ResNet50),APdet and APseg indexes increased by 7.11％and 5.14％,respectively.DIF-Mask RCNN model can effectively detect and segment tumor lesions.It provides important reference value and evaluation basis for computer-aided diagnosis of lung cancer.
关键词
正电子发射断层显像/X线计算机体层成像/实例分割/交互式融合/密集连接/Mask/RCNNKey words
Positron emission tomography/computed tomography/Instance segmentation/Interactive fusion/Dense connection/Mask RCNN引用本文复制引用
基金项目
国家自然科学基金项目(62062003)
宁夏自然科学基金(2023AAC03293)
北方民族大学引进人才科研启动项目(2020KYQD08)
2020年北方民族大学研究生创新项目(YCX21089)
出版年
2024
NETL
NSTL
万方数据