HF-UNet: Learning Hierarchically Inter-Task Relevance in Multi-Task U-Net for Accurate Prostate Segmentation
Nanjing University(南京大学);National Institute of Healthcare Data Science at Nanjing University(南京大学医疗健康数据科学国家研究院);Nanjing Drum Tower Hospital, Nanjing University Medical School(南京大学医学院附属鼓楼医院);Shanghai United Imaging Intelligence Co., Ltd.(上海联影智能科技有限公司);Korea University(高丽大学)
Kelei He, Chunfeng Lian, Bing Zhang, Xin Zhang, Xiaohuan Cao, Dong Nie, Yang Gao, Junfeng Zhang, Dinggang Shen
Accurate segmentation of the prostate is a key step in external beam radiation therapy treatments. In this paper, we tackle the challenging task of prostate segmentation in CT images by a two-stage network with 1) the first stage to fast localize, and 2) the second stage to accurately segment the prostate. To precisely segment the prostate in the second stage, we formulate prostate segmentation into a multi-task learning framework, which includes a main task to segment the prostate, and an auxiliary task to delineate the prostate boundary. Here, the second task is applied to provide additional guidance of unclear prostate boundary in CT images. Besides, the conventional multi-task deep networks typically share most of the parameters (i.e., feature representations) across all tasks, which may limit their data fitting ability, as the specificities of different tasks are inevitably ignored. By contrast, we solve them by a hierarchically-fused U-Net structure, namely HF-UNet. The HF-UNet has two complementary branches for two tasks, with the novel proposed attention-based task consistency learning block to communicate at each level between the two decoding branches. Therefore, HF-UNet endows the ability to learn hierarchically the shared representations for different tasks, and preserve the specificities of learned representations for different tasks simultaneously. We did extensive evaluations of the proposed method on a large planning CT image dataset, including images acquired from 339 patients. The experimental results show HF-UNet outperforms the conventional multi-task network architectures and the state-of-the-art methods.
A Dilated Residual Hierarchically Fashioned Segmentation Framework for Extracting Gleason Tissues and Grading Prostate Cancer from Whole Slide Images
Khalifa University(哈利法大学);Beihang University(北京航空航天大学);University of Louisville(路易斯维尔大学)
前列腺癌(PCa)是男性中第二致命的癌症形式,并且可以通过检查Gleason组织的结构表示来进行临床分级。本文提出了\RV{a new method}分割格里森组织\RV{(patch-wise) in order to grade PCa from the whole slide images (WSI).}此外,所提出的方法包括两个主要贡献:1)一个协同的混合膨胀因子和层次分解的潜在空间表示有效格里森组织提取,和2)一个三层的损失函数,可以惩罚不同的语义分割模型,准确地提取高度相关的模式。除此之外,在包含10,516个完整载玻片扫描(约71.7M块)的大规模PCa数据集上对所提出的框架进行了广泛评估,其中它在提取Gleason组织方面优于最先进的方案3.22%(就平均交叉-联合而言),在PCa进展分级方面优于最先进的方案6.91%(就F1评分而言)。
Deep Learning Based Dominant Index Lesion Segmentation for MR-guided Radiation Therapy of Prostate Cancer
剂量递增放射治疗允许增加对前列腺癌(PCa)的控制,但需要分割显性指数病变(DIL),这促使开发了用于快速、准确和一致分割PCa DIL的自动化方法。我们评估了来自内部训练数据集1(带直肠内线圈的1.5特斯拉GE MR)、外部ProstateX数据集2(3特斯拉Siemens MR)和内部评估者间数据集3(3特斯拉Philips MR)的365例患者500处病变的表观扩散系数(ADC)MRI的5个深度学习网络。这些网络包括:多分辨率剩余连接网络(MRRN)和在训练中使用深度监督(MRRN-DS)、Unet、Unet++、ResUnet和快速全景分割(FPSnet)以及使用平滑标签的快速全景分割(FPSnet-SL)进行正则化的MRRN。通过使用Dice相似系数(DSC)的体积DIL分割准确度和检测准确度(作为病变侵袭性、大小和位置的函数)(数据集1和2)以及相对于两名评定者的准确度(数据集3)评价模型。一般来说,MRRN-DS在测试数据集上比其他方法更准确地分割肿瘤。MRRN-DS在数据集2中显著优于ResUnet(DSC为0.54 vs. 0.44,p<0.001),在数据集3中显著优于Unet++(DSC为0.45 vs. p=0.04)。在数据集2中,FPSnet-SL与MRRN-DS的准确性相似(p = 0.30),但在数据集1(分别为0.60 vs 0.51 [p=0.0…
Performance Deterioration of Deep Learning Models after Clinical Deployment: A Case Study with Auto-segmentation for Definitive Prostate Cancer Radiotherapy
University of Texas Southwestern Medical Center(德克萨斯大学西南医学中心);Southern Methodist University(南卫理公会大学);University of Texas at Arlington(德克萨斯大学阿灵顿分校)
DeepEdit: Deep Editable Learning for Interactive Segmentation of 3D Medical Images
School of Biomedical Engineering & Imaging Sciences, King’s College London(伦敦国王学院生物医学工程与影像科学学院);NVIDIA(英伟达);Semmelweis University(塞麦尔维斯大学);EBATINCA, S.L.(EBATINCA有限公司);Beth Israel Deaconess Medical Center(贝斯以色列女执事医疗中心);Isomics, Inc.(Isomics公司)
Med-NCA: Robust and Lightweight Segmentation with Neural Cellular Automata
Darmstadt University of Technology(达姆施塔特工业大学)
在使用深度学习执行医学图像分割时,访问适当的基础设施至关重要。这一要求使得在资源有限的情况下,如农村地区和危机期间的初级保健设施,很难运行最先进的细分模型。最近新兴的神经元胞自动机(NCA)领域已经表明,局部相互作用的单细胞模型可以在低分辨率输入的图像生成或分割等任务中取得有竞争力的结果。然而,它们受到高VRAM要求和难以达到高分辨率图像的收敛的限制。为了克服这些限制,我们提出了Med-NCA,一个用于高分辨率图像分割的端到端NCA训练管道。我们的方法遵循两个步骤的过程。全局知识首先在缩小的图像中的细胞之间进行通信。之后,执行基于块的分割。我们提出的Med-NCA在海马和前列腺分割方面分别比经典的UNet好2%和3%,同时也小了500倍。我们还表明,Med-NCA是通过设计不变的图像尺度,形状和平移,经历只有轻微的性能下降,即使有很强的转变,是强大的MRI采集伪影。Med-NCA甚至可以在Raspberry Pi B+上实现高分辨率医学图像分割,Raspberry Pi B+可以说是能够运行PyTorch的最小设备,并且可以由标准电源供电。