我们引入了贝叶斯QuickNAT,用于MRI T1扫描的全脑分割的自动化质量控制。在贝叶斯完全卷积神经网络的旁边,我们还提出了分段不确定性的固有度量,允许每个大脑结构的质量控制。为了估计模型不确定性,我们遵循贝叶斯方法,其中,通过在测试时保持丢失层活动来生成来自后验分布的蒙特卡罗(MC)样本。 MC样本上的熵提供了体素模型不确定性图,而对MC预测的期望提供了最终分割。除了体素不确定性之外,我们还在质量控制的分割中引入了四个量度结构的结构不确定性。对包含不同年龄范围,病理学和成像伪影的四个样本外数据集进行实验。所提出的结构不确定度量与通过manualannotation估计的Dice得分高度相关,因此呈现了分割质量的固有度量。特别地,所有MC样本上的交叉结合是Dice得分的合适代理。除了质量控制atscan级别,我们建议将结构方面的不确定性作为一种信心度量来对大型数据存储库进行可靠的组分析。我们设想引入的不确定性度量将有助于评估基于自动深度学习的分割方法的保真度。用于大规模人口研究,因为它们可以在处理大型数据存储库时实现自动化质量控制和组分析。
translated by 谷歌翻译
在医学成像中使用深度学习已经在研究界获得了巨大的发展。在临床环境中缓慢采用这些系统的一个原因是它们是复杂的,不透明的并且往往会无声地失败。在医学成像领域之外,机器学习社区最近提出了几种用于量化模型不确定性的技术(即〜模型知道何时它失败了)。这在实际环境中很重要,因为我们可以将此类情况转交给人工进行人工检查或更正。在本文中,我们的目标是将这些最近的结果用于估计不确定性,从而在基于深度学习的分割中承担两个重要的输出。第一个产生空间不确定性的地图,临床医生可以从中观察系统认为它失败的位置和原因。第二个是量化故障的图像级预测,这对于隔离特定情况并从自动化管道中移除它们非常有用。我们还表明,关于空间不确定性的推理,即第一个输出,是生成分割质量预测的有用中间表示,第二个输出。我们提出了一种用于产生这些不确定性测量的两阶段架构,它可以适应任何基于深度学习的医学分割管道。
translated by 谷歌翻译
心脏图像分割是用于生成心脏的个性化模型和用于量化心脏性能参数的关键过程。已经提出了几种卷积神经网络(CNN)架构来从心脏电影MR图像分割心室。在这里,我们提出了一个基于多任务学习(MTL)的心脏MR图像分割正则化框架。训练网络以执行语义分割的主要任务,以及像素方式距离图回归的同时辅助任务。所提出的距离图正则化器是一种解码器网络,其加入现有CNN架构的瓶颈层,便于网络学习强大的全局特征。在调整后移除正则化块,使得原始数量的网络参数不会改变。 Weshow所提出的正则化方法在两个公开可用的心脏电影MRI数据集上相对于相应的现有CNN架构改善了二进制和多类分割性能,获得的平均骰子系数为0.84 $ \ pm $ 0.03和0.91 $ \ pm $ 0.04 ,分别。此外,我们还展示了改进的距离图正则化网络在跨数据集分割上的泛化性能,显示平均骰子系数从0.57 $ \ pm $ 0.28提高41%至0.80 $ \ pm $ 0.13。
translated by 谷歌翻译
我们引入了基于贝叶斯完全卷积神经网络的有效质量控制脑分割的固有措施,使用模型不确定性。在测试时使用压差可以有效地生成来自后验分布的蒙特卡罗样本。基于这些样本,我们在体素方面的不确定性图谱旁边引入了三个度量结构方面的不确定性。然后,我们将这些结构明确不确定性纳入群体分析中作为观察信心的度量。我们的结果表明,度量与分割准确性高度相关,因此提供了分割质量的固有度量。此外,具有不确定性的群体分析导致影响大小更接近手工注释。引入的不确定性指标不仅在转化为临床实践时非常有用,而且在处理大型数据存储库时还提供自动质量控制和组分析。
translated by 谷歌翻译
尽管对自动方法有很高的要求,但目前最先进的深度学习分割方法还没有广泛进入临床环境。一个重要的原因是缺乏可靠性导致的模型缺乏可靠性,并且经常在本地产生医学专家不会做出的解剖学上令人难以置信的结果。本文提出了一种基于(贝叶斯)扩张卷积网络(DCNN)的自动图像分割方法,该方法为手头的输入图像生成分割掩模和空间不确定性图。在MICCAI 2017的100个心脏2D MR扫描中,使用左心室(LV)腔,右心室(RV)心内膜和心肌(Myo)在舒张末期(ED)和收缩末期(ES)的分割来训练和评估该方法。挑战(ACDC)。结合分割和不确定性图并采用人在环设置,我们提供的证据表明,所获得的分割高度不确定的图像区域可以完全覆盖不正确分割的区域。融合信息可以表现为增加分割性能。我们的结果表明,我们可以使用DCNN以低计算量获得有价值的空间不确定性图。
translated by 谷歌翻译
脊髓经常受到多发性硬化(MS)患者的萎缩和/或损伤的影响。来自MRI数据的脊髓和病变的分割提供了损伤的测量,这是MS的诊断,预后和纵向监测的关键标准。自动化此操作可消除评估者之间的差异,并提高大吞吐量分析管道的效率。由于与采集参数和图像伪像相关的大的变异性,跨多位点脊髓数据的稳健且可靠的分割是具有挑战性的。本研究的目的是开发一种全自动框架,对图像参数和临床条件的可变性具有鲁棒性,可用于从传统MRI数据中分割脊髓和髓内MS病变。在这项多地点研究中纳入了1,042名受试者(459名健康对照,471名MS患者和112名其他脊柱病变)的扫描(n = 30)。数据跨越三个对比(T1-,T2-和T2 * - 加权),共计1,943卷。所提出的脐带病变自动分割方法基于两个卷积神经网络(CNN)的序列。为了处理非常小比例的脊髓和/或病变体素与体积的其余部分相比,具有2D扩张卷曲的第一CNN检测脊髓中心线,接着是第二CNN,其具有分割脊髓和/或脊髓的3D卷曲。与手动分割相比,我们基于CNN的方法显示中位数骰子为95%而PropSeg为88%,这是一种先进的脊髓分割方法。关于MS数据的病变分割,我们的框架提供了60%的骰子,-15%的相对体积差异,以及83%和77%的病变检测灵敏度和精确度。建议的框架是开源的,可以在脊髓工具箱中随时使用。
translated by 谷歌翻译
现代神经网络的不确定性估计在计算的预测旁边提供了额外的信息,因此期望提高对基础模型的理解。对于安全关键的计算机辅助应用,例如神经外科干预和放射治疗计划,可靠的不确定性尤其令人感兴趣。我们提出了一个不确定性驱动的健全性检查,以确定需要特殊专家审查的分割结果。我们的方法使用一个卷积神经网络,并根据蒙特卡洛辍学的原则计算不确定性估计。我们评估了所提出的方法在具有30个术后脑肿瘤图像的临床数据集上的性能。方法可以准确地分割高度不均匀的切除腔(Dicecoefficients 0.792 $ \ pm $ 0.154)。此外,所提出的健全性检查能够检测最差的分割和四个异常值中的三个。结果突出了使用来自模型参数不确定性的附加信息来验证陡峭学习模型的分割性能的潜力。
translated by 谷歌翻译
光学相干断层扫描(OCT)通常用于分析视网膜层以评估眼部疾病。在本文中,我们提出了一种基于贝叶斯深度学习的视觉层分割和不确定性量化方法。我们的方法不仅执行视网膜层的端到端分割,而且还给出分割输出的像素方式不确定性度量。生成的不确定性图可用于识别在下游分析中有用的错误分割的图像区域。我们在从15个受试者(OCT体积)获得的1487个图像的数据集上验证了我们的方法,并将其与不考虑不确定性的最新分割算法进行了比较。所提出的基于不确定性的分割方法导致相当或改进的性能,并且最重要的是对抗噪声更强。
translated by 谷歌翻译
白质病变和深灰质结构的分割是多发性硬化中磁共振成像量化的重要任务。通常这些任务是分开执行的:在本文中,我们提出了一个基于CNN的分段解决方案,用于快速,可靠地将多模态MR图像分割为病变类和健康的灰色和白质结构。与先前的方法相比,我们在骰子系数和病变特异性和敏感性方面显示出显着的,统计学上显着的改善,并且在人类内部评估者范围内与个体人类评价者协商。该方法是针对从单个中心收集的数据进行训练的:尽管如此,它对来自训练数据集中未表示的中心,扫描仪和场强的数据表现良好。一项回顾性研究发现,分类器成功识别出人类遗漏的病变。损伤标签由人类评估者提供,而其他脑结构(包括脑脊液,皮质灰质,皮质白质,小脑,扁桃体,海马,皮质下GM结构和脉络膜复合体)的弱标签由Freesurfer 5.3提供。这些结构的分割不仅与Freesurfer 5.3相当,而且与FSL-First和Freesurfer 6.1相比也很好。
translated by 谷歌翻译
从不同医学成像模式同时分割多个器官是一项至关重要的任务,因为它可用于计算机辅助诊断,计算机辅助手术和治疗计划。由于深度学习的进步,已经成功地为此目的引入了几个用于医学图像分割的深度神经网络。在本文中,我们专注于学习标注体素的深层多器官分割网络。特别是,我们研究了损失函数的关键选择,以便处理困扰学习模型的输入和输出的臭名昭着的不平衡问题。输入不平衡指的是输入训练样本中的类不平衡(即,嵌入在背景体素的多个中的小前景对象,以及不同大小的器官)。 outputibalance指的是推理模型的误报和假阴性之间的不平衡。为了解决这两种不平衡的培训和推理问题,我们引入了一种基于新课程学习的损失函数。具体而言,我们利用Dice相似性系数来确定参数,使其保持在较差的局部最小值,同时通过使用交叉熵项惩罚假阳性/阴性来逐渐学习更好的模型参数。我们在三个数据集上评估了所提出的损失函数:具有5个靶器官的全身正电子发射断层扫描(PET)扫描,磁共振成像(MRI)前列腺扫描,以及具有单个靶器官即左心室的超声心动图。我们表明,具有所提出的一体化损失函数的简单网络架构可以胜过最先进的方法,并且当使用我们提出的损失时,可以改进竞争方法的结果。
translated by 谷歌翻译
最近,深度学习(DL)网络在各种公共医学图像挑战数据集[3,11,16]上表现优于其他分割方法,特别是对于大型病理学。然而,在诸如多发性硬化症(MS)等疾病的背景下,监测MRI序列上可见的所有局灶性病变,甚至是非常小的病灶,对于疾病分期,预后和评估治疗功效是必不可少的。此外,产生确定性的输出阻碍了DL在临床常规中的应用。预测的不确定性估计将允许临床医生随后进行修订。我们在医学图像中用于病变检测和分割的深度网络的背景下,基于蒙特卡洛(MC)辍学[4]首次探索多个不确定性估计。具体来说,我们开发了一个3D MSlesion分割CNN,增加了基于MC丢失的四种不同的基于体素的不确定性测量。我们通过专有的,大规模的,多站点,多扫描仪,临床MS数据集训练网络,并通过从检测到的病变内的体素不确定性中积累证据来计算病变不确定性。我们通过基于不确定性选择操作点来分析基于体素的分割和病变水平检测的性能。经验证据表明,不确定性措施始终允许我们选择优越的操作点,仅使用网络的S形模型输出作为概率。
translated by 谷歌翻译
Brain extraction from magnetic resonance imaging (MRI) is crucial for many neuroimaging workflows. Current methods demonstrate good results on non-enhanced T1-weighted images, but struggle when confronted with other modalities and pathologically altered tissue. In this paper we present a 3D convolutional deep learning architecture to address these shortcomings. In contrast to existing methods, we are not limited to non-enhanced T1w images. When trained appropriately, our approach handles an arbitrary number of modalities including contrast-enhanced scans. Its applicability to MRI data, comprising four channels: non-enhanced and contrast-enhanced T1w, T2w and FLAIR contrasts, is demonstrated on a challenging clinical data set containing brain tumors (N = 53), where our approach significantly outperforms six commonly used tools with a mean Dice score of 95.19. Further, the proposed method at least matches state-of-the-art performance as demonstrated on three publicly available data sets: IBSR, LPBA40 and OASIS, totaling N = 135 volumes. For the IBSR (96.32) and LPBA40 (96.96) data set the convolutional neuronal network (CNN) obtains the highest average Dice scores, albeit not being significantly different from the second best performing method. For the OASIS data the second best Dice (95.02) results are achieved, with no statistical difference in comparison to the best performing tool. For all data sets the highest average specificity measures are evaluated, whereas the sensitivity displays about average results. Adjusting the cutoff threshold for generating the binary masks from the CNN's probability output can be used to increase the sensitivity of the method. Of course, this comes at the cost of a decreased specificity and has to be decided application specific. Using an optimized GPU implementation predictions can be achieved in less than one minute. The proposed method may prove useful for large-scale studies and clinical trials.
translated by 谷歌翻译
完全卷积的深度神经网络被认为是快速和精确的框架,在图像分割方面具有巨大的潜力。当数据不平衡时,利用这种网络的主要挑战之一就会增加,这在诸如病变分割的许多医学成像应用中是常见的,其中病变类体素的数量通常比非病变体素低得多。具有不平衡数据的训练网络可以进行具有高精度和低召回率的预测,严重偏向非病变类别,这在医学应用中是特别不希望的,其中假阴性实际上比假阳性更重要。已经提出了各种方法来解决该问题,包括两步训练,采样加权,平衡采样和相似性损失函数。在本文中,我们开发了一个具有非对称损失函数的拼接三维密集连接网络,其中我们使用大的重叠图像补丁进行内在的内部数据增强,补丁选择算法和基于B样条加权软投票的补丁预测融合策略inaccount补丁边界预测的不确定性。我们将此方法应用于基于MSSEG 2016和ISBI 2015挑战的病变分割,其中我们的平均Dice相似系数分别为69.9%和65.74%。除了提议的损失,我们训练我们的网络具有局部和广义的骰子丢失功能。使用非对称相似性损失层和我们的3D贴片预测融合,在测试中实现了$ F_1 $和$ F_2 $得分以及APR曲线的显着改善。基于$ F_ \ beta $得分的非对称相似性损失推广了Dice相似系数,并且可以有效地与开发的补丁策略一起用于训练完全卷积深度神经网络以进行高度不平衡图像分割。
translated by 谷歌翻译
Deep convolutional neural networks are powerful tools for learning visual representations from images. However, designing efficient deep architectures to analyse volumetric medical images remains challenging. This work investigates efficient and flexible elements of modern convolutional networks such as dilated convolution and residual connection. With these essential building blocks, we propose a high-resolution, compact convolutional network for volumetric image segmentation. To illustrate its efficiency of learning 3D representation from large-scale image data, the proposed network is validated with the challenging task of parcellating 155 neuroanatomical structures from brain MR images. Our experiments show that the proposed network architecture compares favourably with state-of-the-art volumetric segmentation networks while being an order of magnitude more compact. We consider the brain par-cellation task as a pretext task for volumetric image segmentation; our trained network potentially provides a good starting point for transfer learning. Additionally, we show the feasibility of voxel-level uncertainty estimation using a sampling approximation through dropout.
translated by 谷歌翻译
在本文中,我们展示了训练有素的U-Net在BraTS 2018挑战中的有效性。这项努力特别有趣,因为研究人员目前正在通过旨在提高分割性能的架构修改相互优化。 Weinstead专注于训练过程,认为训练有素的U-Net是难以击败的,并打算在今年的BraTS挑战中积极参与这一假设。我们的基线U-Net只有少量修改,并且经过大修补程序和骰子丢失功能的训练,可以在BraTS2018验证数据上获得有竞争力的骰子评分。通过结合基于区域的训练,额外的训练数据和简单的后处理技术,我们得到的骰子得分分别为81.01,90.83和85.44以及哈斯多夫距离(第95百分位数)为2.54,4.97和7。
translated by 谷歌翻译
自动医学图像分割,特别是使用深度学习,在语义分割任务中具有出色的表现。然而,这些方法很少量化它们的不确定性,这可能导致错误的下游分析。在这项工作中,我们建议使用贝叶斯神经网络来量化语义分割领域内的不确定性。我们还提出了一种方法,将体素分割不确定性转换为体积不确定性,并校准衍生测量的置信区间的准确性和可靠性。当应用于肿瘤体积估计应用时,我们证明通过使用这种不确定性建模,可以使用深度学习系统来报告具有良好校准误差条的体积估计,使其更安全地用于临床。我们还表明,不确定性估计推断出看不见的数据,并且在存在人工噪声时置信区间是稳健的。这可用于提供质量控制和质量保证的形式,并可允许在诊所中进一步使用深度学习工具。
translated by 谷歌翻译
Deep learning techniques are being rapidly applied to medical imaging tasks-from organ and lesion segmenta-tion to tissue and tumor classification. These techniques are becoming the leading algorithmic approaches to solve inherently difficult image processing tasks. Currently, the most critical requirement for successful implementation lies in the need for relatively large datasets that can be used for training the deep learning networks. Based on our initial studies of MR imaging examinations of the kidneys of patients affected by polycystic kidney disease (PKD), we have generated a unique database of imaging data and corresponding reference standard segmentations of polycystic kidneys. In the study of PKD, segmentation of the kidneys is needed in order to measure total kidney volume (TKV). Automated methods to segment the kidneys and measure TKV are needed to increase measurement throughput and alleviate the inherent variability of human-derived measurements. We hypothesize that deep learning techniques can be leveraged to perform fast, accurate, reproducible, and fully automated segmentation of polycystic kidneys. Here, we describe a fully automated approach for segmenting PKD kidneys within MR images that simulates a multi-observer approach in order to create an accurate and robust method for the task of segmentation and computation of TKV for PKD patients. A total of 2000 cases were used for training and validation, and 400 cases were used for testing. The multi-observer ensemble method had mean ± SD percent volume difference of 0.68 ± 2.2% compared with the reference standard segmentations. The complete framework performs fully automated segmentation at a level comparable with interobserver variability and could be considered as a replacement for the task of segmentation of PKD kidneys by a human.
translated by 谷歌翻译
Gray matter (GM) tissue changes have been associated with a wide range of neurological disorders and were recently found relevant as a biomarker for disability in amyotrophic lateral sclerosis. The ability to automatically segment the GM is, therefore, an important task for modern studies of the spinal cord. In this work, we devise a modern, simple and end-to-end fully-automated human spinal cord gray matter segmentation method using Deep Learning, that works both on in vivo and ex vivo MRI acquisitions. We evaluate our method against six independently developed methods on a GM segmentation challenge. We report state-of-the-art results in 8 out of 10 evaluation metrics as well as major network parameter reduction when compared to the traditional medical imaging architectures such as U-Nets. Gray matter (GM) and white matter (WM) tissue changes in the spinal cord (SC) have been linked to a large spectrum of neurological disorders 1. For example, using magnetic resonance imaging (MRI), the involvement of the spinal cord gray matter (SCGM) area in multiple sclerosis (MS) was found to be the strongest correlate of disability in multivariate models including brain GM and WM volumes, FLAIR lesion load, T1-lesion load, SCWM area, number of spinal cord T2 lesions, age, sex and disease duration 2. Another study showed SCGM atrophy to be a biomarker for predicting disability in amyotrophic lateral sclerosis 3. The ability to automatically assess and characterize these changes is, therefore, an important step 4 in the modern pipeline to study both the in vivo and ex vivo SC. The segmentation outcome can also be used for co-registration and spatial normalization to a common space. Moreover, the fully-automated segmentation is useful for longitudinal studies, where the delineation of gray matter is time consuming 4. While recent cervical cord cross-sectional area (CSA) segmentation methods have achieved near-human performance 5 , the accurate segmentation of the GM remains a challenge 6. The main properties that make the GM area difficult to segment are: inconsistent intensities of the surrounding tissues, image artifacts and pathology-induced changes in the image contrast 4. Additional factors also contribute to the complexity of the GM segmentation task, such as lack of standardized datasets, differences in MRI acquisition protocols, different pixel sizes, different methods to acquire gold standard segmentations and different performance metrics to assess segmentation results 6. Figure 1 features several examples of axial MRI acquired at different centers, demonstrating image variability due variable image acquisition systems and protocols. Despite these difficulties, there have been major improvements in acquisition and analysis methods in recent years, making it possible to obtain reliable GM segmentations. From the acquisition standpoint, the advances in coil sensitivity 7 , multi-echo gradient echo sequences 8 , and phase-sensitive inversion recovery se
translated by 谷歌翻译
基于深度学习的图像分割方法的最新进展已经具有人类精确度的实时性能。然而,偶尔最好的方法由于低图像质量,伪像或黑盒算法的意外行为而失败。能够在没有基础事实的情况下预测分割质量在临床实践中是至关重要的,但在大规模研究中也是如此,以避免在随后的分析中包含无效数据。在这项工作中,我们提出了两种使用深度学习进行心血管MR分割的实时自动质量控制方法。首先,在12,880个样本上对一个神经网络进行润湿,以便根据每个案例预测Dice相似系数(DSC)。我们报告1,610个测试样本的平均误差(MAE)为0.03,二元分类精度为97%,这反映了低质量和高质量的分割。其次,在没有手动注释数据可用的情况下,我们训练网络来预测通过反向测试策略获得的估计质量的DSC分数。对于这种情况,Wereport的MAE = 0.14和91%二进制分类精度。实时获得预测,当与实时分割方法结合时,能够在患者仍在扫描仪中时获得关于获取的扫描是否是可分析的即时反馈。这进一步使得优化图像采集的新应用朝向最佳可能的分析结果。
translated by 谷歌翻译
Stroke is an acute cerebral vascular disease, which is likely to cause long-term disabilities and death. Acute ischemic lesions occur in most stroke patients. These lesions are treatable under accurate diagnosis and treatments. Although diffusion-weighted MR imaging (DWI) is sensitive to these lesions, localizing and quantifying them manually is costly and challenging for clinicians. In this paper, we propose a novel framework to automatically segment stroke lesions in DWI. Our framework consists of two convolutional neural networks (CNNs): one is an ensemble of two DeconvNets (Noh et al., 2015), which is the EDD Net; the second CNN is the multi-scale convolutional label evaluation net (MUSCLE Net), which aims to evaluate the lesions detected by the EDD Net in order to remove potential false positives. To the best of our knowledge, it is the first attempt to solve this problem and using both CNNs achieves very good results. Furthermore, we study the network architectures and key configurations in detail to ensure the best performance. It is validated on a large dataset comprising clinical acquired DW images from 741 subjects. A mean accuracy of Dice coefficient obtained is 0.67 in total. The mean Dice scores based on subjects with only small and large lesions are 0.61 and 0.83, respectively. The lesion detection rate achieved is 0.94.
translated by 谷歌翻译