FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W4368360866> ?p ?o ?g. }

• W4368360866 endingPage "201" @default.
• W4368360866 startingPage "193" @default.
• W4368360866 abstract "When applying deep learning algorithms to magnetic resonance (MR) image segmentation, a large number of annotated images are required as data support. However, the specificity of MR images makes it difficult and costly to acquire large amounts of annotated image data. To reduce the dependence of MR image segmentation on a large amount of annotated data, this paper proposes a meta-learning U-shaped network (Meta-UNet) for few-shot MR image segmentation. Meta-UNet can use a small amount of annotated image data to complete the task of MR image segmentation and obtain good segmentation results. Meta-UNet improves U-Net by introducing dilated convolution, which can increase the receptive field of the model to improve the sensitivity to targets of different scales. We introduce the attention mechanism to improve the adaptability of the model to different scales. We introduce the meta-learning mechanism, and employ a composite loss function for well-supervised and effective bootstrapping of model training. We use the proposed Meta-UNet model to train on different segmentation tasks, and then use the trained model to evaluate on a new segmentation task, where the Meta-UNet model achieves high-precision segmentation of target images. Meta-UNet has a certain improvement in mean Dice similarity coefficient (DSC) compared with voxel morph network (VoxelMorph), data augmentation using learned transformations (DataAug) and label transfer network (LT-Net). Experiments show that the proposed method can effectively perform MR image segmentation using a small number of samples. It provides a reliable aid for clinical diagnosis and treatment.将深度学习算法应用于核磁共振（MR）图像分割时，必需以大量经标注后图像作为训练集的数据支撑。然而，MR图像的特殊性导致采集大量的图像数据较困难，制作大量的标注数据成本高。为降低MR图像分割对大量标注数据的依赖，本文提出了一种用于小样本MR图像分割的元U型网络(Meta-UNet)，能够利用少量的图像标注数据完成MR图像分割任务，并获得良好的分割结果。其具体操作为：通过引入空洞卷积对U型网络（U-Net）进行改进，增加网络模型感受野从而提高模型对不同尺度目标的灵敏度；通过引入注意力机制提高模型对不同尺度目标的适应性；通过引入元学习机制，并采用复合损失函数对模型训练进行良好的监督和有效的引导。本文利用提出的Meta-UNet模型，在不同分割任务上进行训练，然后用训练好的模型在全新的分割任务上进行评估，实现了目标图像的高精度分割。新的分割方法比起常用的无监督医学图像配准分割方法——体素变形网络（VoxelMorph）、数据增强医学图像分割方法——转换学习数据增强模型（DataAug）和基于标签转移的医学图像分割方法——标签转移网络（LT-Net）三种模型平均戴斯相似性系数（DSC）有一定提高。实验结果显示，本文所提方法利用少量样本即可有效地进行MR图像分割，今后可为临床诊断和治疗提供可靠辅助。." @default.
• W4368360866 created "2023-05-05" @default.
• W4368360866 creator A5024764114 @default.
• W4368360866 creator A5036885388 @default.
• W4368360866 creator A5042385997 @default.
• W4368360866 date "2023-04-25" @default.
• W4368360866 modified "2023-09-27" @default.
• W4368360866 title "[A meta-learning based method for segmentation of few-shot magnetic resonance images]." @default.
• W4368360866 cites W1969094805 @default.
• W4368360866 cites W3097812605 @default.
• W4368360866 cites W3104164805 @default.
• W4368360866 cites W3140070796 @default.
• W4368360866 cites W3163842339 @default.
• W4368360866 cites W4229459726 @default.
• W4368360866 doi "https://doi.org/10.7507/1001-5515.202208004" @default.
• W4368360866 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/37139748" @default.
• W4368360866 hasPublicationYear "2023" @default.
• W4368360866 type Work @default.
• W4368360866 citedByCount "0" @default.
• W4368360866 crossrefType "journal-article" @default.
• W4368360866 hasAuthorship W4368360866A5024764114 @default.
• W4368360866 hasAuthorship W4368360866A5036885388 @default.
• W4368360866 hasAuthorship W4368360866A5042385997 @default.
• W4368360866 hasConcept C103278499 @default.
• W4368360866 hasConcept C108583219 @default.
• W4368360866 hasConcept C115961682 @default.
• W4368360866 hasConcept C124504099 @default.
• W4368360866 hasConcept C153180895 @default.
• W4368360866 hasConcept C154945302 @default.
• W4368360866 hasConcept C22019652 @default.
• W4368360866 hasConcept C25694479 @default.
• W4368360866 hasConcept C31972630 @default.
• W4368360866 hasConcept C41008148 @default.
• W4368360866 hasConcept C50644808 @default.
• W4368360866 hasConcept C65885262 @default.
• W4368360866 hasConcept C81363708 @default.
• W4368360866 hasConcept C89600930 @default.
• W4368360866 hasConceptScore W4368360866C103278499 @default.
• W4368360866 hasConceptScore W4368360866C108583219 @default.
• W4368360866 hasConceptScore W4368360866C115961682 @default.
• W4368360866 hasConceptScore W4368360866C124504099 @default.
• W4368360866 hasConceptScore W4368360866C153180895 @default.
• W4368360866 hasConceptScore W4368360866C154945302 @default.
• W4368360866 hasConceptScore W4368360866C22019652 @default.
• W4368360866 hasConceptScore W4368360866C25694479 @default.
• W4368360866 hasConceptScore W4368360866C31972630 @default.
• W4368360866 hasConceptScore W4368360866C41008148 @default.
• W4368360866 hasConceptScore W4368360866C50644808 @default.
• W4368360866 hasConceptScore W4368360866C65885262 @default.
• W4368360866 hasConceptScore W4368360866C81363708 @default.
• W4368360866 hasConceptScore W4368360866C89600930 @default.
• W4368360866 hasIssue "2" @default.
• W4368360866 hasLocation W43683608661 @default.
• W4368360866 hasOpenAccess W4368360866 @default.
• W4368360866 hasPrimaryLocation W43683608661 @default.
• W4368360866 hasRelatedWork W1507266234 @default.
• W4368360866 hasRelatedWork W1669643531 @default.
• W4368360866 hasRelatedWork W2069711651 @default.
• W4368360866 hasRelatedWork W2117664411 @default.
• W4368360866 hasRelatedWork W2117933325 @default.
• W4368360866 hasRelatedWork W2549936415 @default.
• W4368360866 hasRelatedWork W2558375057 @default.
• W4368360866 hasRelatedWork W2739874619 @default.
• W4368360866 hasRelatedWork W2897195263 @default.
• W4368360866 hasRelatedWork W1967061043 @default.
• W4368360866 hasVolume "40" @default.
• W4368360866 isParatext "false" @default.
• W4368360866 isRetracted "false" @default.
• W4368360866 workType "article" @default.