乳腺肿瘤超声图像识别模式分类方法的比较研究

目的利用乳腺肿瘤超声图像良恶性的不同特征,借助于模式分类方法对乳腺肿瘤良恶性进行识别,作为医生的计算机辅助诊断。方法本文研究基于乳腺肿瘤超声图像的原始特征参数已提取情况下,采用顺序前进搜索方法获得最优特征矢量,然后利用支撑矢量机、贝叶斯分类器、BP网络和Fisher线性判别器四种模式识别方法分别对乳腺肿瘤良恶性进行识别。结果基于200例病例随机划分为训练集100例和测试集100例进行测试,支撑矢量机、贝叶斯分类器、BP网络和Fisher线性判别器的Accuracy分别为0.960,0.940,0.932±0.013,0.930。结论支撑矢量机的分类性能优于其它分类器,能有效地对超声图像乳腺肿瘤进行良恶性识别。     

边界增强像素级多模态声像图融合方法评估良、恶性乳腺肿瘤

目的 观察以边界增强多模态乳腺声像图像素级特征融合方法评估良、恶性乳腺肿瘤性质的价值。方法 基于乳腺肿瘤B型声像图提取边界增强图像,于超声弹性复合声像图中提取纯弹性信息图像。对多模态乳腺肿瘤声像图进行像素级特征融合,形成边界特征增强的融合图像,再以卷积神经网络(CNN)进行分类;评估融合方法分类良、恶性乳腺肿瘤的性能,并与单模态方法、特征级融合方法、无边界增强像素级图像融合方法及其他CNN模型进行对比。结果 边界增强像素级特征融合方法有助于CNN提取乳腺肿瘤特征,分类良、恶性乳腺性能最佳,其分类准确率为85.71%,特异度为85.49%,敏感度为86.16%,模型稳定。结论 边界特征增强像素级多模态声像图融合方法可用于判断良、恶性乳腺肿瘤。     

灰阶超声鉴别诊断良、恶性乳腺肿瘤

目的探讨灰阶超声鉴别良、恶性乳腺肿瘤的价值。方法利用改进的Level Set变分模型对126例乳腺肿瘤的超声图像进行分割,提取肿瘤边界,分别计算16个形态特征参数,结合特征参数间的相关性及部分特征参数性质确定特征向量组合,最后用模糊C-均值方法鉴别乳腺肿瘤的良、恶性。结果 126例中,恶性肿瘤50例,良性肿瘤76例。通过Level Set模型得到了较好的分割良、恶性的准确率达80.95%(102/126),其敏感度、特异度、阳性预测值和阴性预测值分别为80.00%(40/50)、81.58%(62/76)、74.07%(40/54)和86.11%(62/72)。结论良、恶性乳腺肿瘤在形态上有较大差异,灰阶超声可有效鉴别乳腺肿瘤的性质。     

DFY-Ⅱ型超声图像定量分析仪对乳腺良恶性肿瘤的鉴别诊断

目的 研究DFY-Ⅱ型超声图像定量分析仪对乳腺良恶性肿瘤的量化诊断方法。方法 采集手术确诊的良恶性乳腺肿瘤各20例的原始超声图像,使用DFY-II型超声图像定量分析仪提取两组图像的纹理特征和灰度特征参数,对比分析两组问各参数的差异。结果 良恶性乳腺肿瘤患者超声图像的熵、平均灰度、平均声强、扭曲度、边缘不规则度及纵横比参数比较,差异有统计学意义（P〈0．05）。结论 DFY-Ⅱ型超声图像定量分析仪可量化乳腺肿瘤的超声图像特征,熵、平均灰度、平均声强、扭曲度、边缘不规则度及纵横比对乳腺良恶性肿块的超声诊断量化有一定参考价值。     

Snake模型在乳腺肿瘤超声图像处理中的运用

目的对乳腺超声图像中的肿瘤进行边缘提取。方法鉴于医学超声图像的信噪比较低,用经典的边缘提取算法无法得到较好的结果,因此,Snake模型作为一种基于高层信息的有效目标轮廓提取算法而引起广泛的关注。在原始的Snake模型的基础上,本文针对超声图像的特点对它进行了一些改进。结果从上海第六人民医院采集到乳腺超声图像15幅。在进行了灰度分割、形态滤波等一系列预处理后,将改进后的Snake模型运用到边缘提取中来,并在这15幅图像中得到了比较好的分割结果。结论改进后的Snake模型可以将乳腺超声图像中肿瘤的边缘较好地提取出来,为乳腺肿瘤计算机辅助诊断提供了重要依据。     

计算机辅助鉴别乳腺良、恶性肿块的临床价值

目的 提出一种基于灰度级二维直方图的计算机辅助分割算法,对乳腺高频超声图像中的肿块进行自动识别和分割处理,旨在提高乳腺良、恶性肿块的检出准确率.方法 采集100个乳腺肿块二维超声图像共466张(原片),应用计算机软件对原片进行分割处理,得到分割后图片.超声医师采用双盲法分别根据原片和分割后图片中的超声征象进行良、恶性判断,运用受试者工作曲线(ROC曲线)计算曲线下面积(A),比较前后两次诊断结果 ,分析图片处理前后诊断结果 的差异性.结果 处理后的图片中乳腺肿块的边缘、钙化等信息明显突出.超声医师对良、恶性肿块的确诊率明显提高.当特异性为74.31%时,诊断敏感性由基于原片的70.32%提高到图片分割后的90.52%.ROC曲线下面积由分割前的80.8%上升到90.5%,差异有统计学意义(P<0.01).结论 此分割算法能明显优化乳腺肿块的边缘信息,较好地突显肿块内微钙化,在一定程度上降低漏诊率和误诊率,提高乳腺良恶性肿块的确诊率.     

超声乳腺肿瘤图像的边缘提取

目的探求乳腺肿瘤超声图像的边缘提取。方法广义梯度矢量流Snake模型已经成功地用于噪声相对比较小的CT、MRI等医学图像,然而乳腺肿瘤超声图像对比度低,斑点噪声大,很难将该模型直接应用于乳腺肿瘤超声图像。本文针对乳腺肿瘤超声图像的特点如图像对比度低,斑点噪声大,部分边缘缺失,肿瘤内部微细结构分布复杂(如血管,钙化灶等),特别恶性肿瘤还具有复杂形状等,采用相应的图像处理技术如非线性各向异性扩散滤除斑点噪声,形态学滤波器平滑图像,直方图均衡化提高图像的对比度,最后将该模型引入到乳腺肿瘤超声图像边缘提取。结果实验对158例乳腺肿瘤超声图像进行边缘提取,定量和定性分析均获得满意的结果。结论本文方法可以有效地用于超声乳腺肿瘤图像的边缘提取。     

乳腺肿瘤灰阶超声造影灌注模式研究

目的 探讨良恶性乳腺肿瘤灰阶超声造影的灌注模式.方法 142例经手术病理证实良恶性的乳腺肿瘤患者术前行灰阶超声造影,评估灌注模式各项观察指标,包括灌注运动形式、有无灌注缺损、有无边缘血管样灌注、灌注强度、造影后边界是否清晰等.结果 良恶性乳腺肿瘤在灌注运动模式和造影后边界上均无统计学意义(P＞0.05).乳腺癌与良性乳腺肿瘤在灌注强度(P=0.000)、有无灌注缺损(P=0.025)、有无边缘放射状灌注(P=0.000)等灌注模式上均有统计学意义.结论 超声造影肿瘤的灌注模式能反映乳腺良恶性肿瘤的血管特征.     

基于纹理分析和双边滤波方法对加标记心脏核磁共振图像分割

背景:加标记心脏核磁共振成像方式提供了左心室内外心膜的边缘信息,该边缘信息可由分割图像得到.但是,所引入的标记线加大了这类图像边界分割的凼难.目的:针对目前在加标记心脏核磁共振图像中对左心室分割困难的问题,提出了一种新的自动分割的方法.方法;首先,使用全局直方图规定化方法增强标记和非标记区域的对比度;然后,利用一种简单的纹理分析方法区分血流充盈的心腔(非纹理)区域和加标记心肌(纹理)区域;再应用双边滤波在保持边界的同时滤掉图像的伪影;最后,用GVF-snake模型自动提取左心室图像的边界.结果与结论:提出了一种简单的纹理分析方法来移除标记线:用局部窗口中的最大灰度值与最小灰度值之差来代替原象素点灰度值,再运用双边滤波滤除图像伪影并保持边界,最后应用 GVF-snake模型实现了左心室边界的有效提取.实验结果显示,该方法能够较好地提取部分加标记心脏核磁共振图像中血流充盈区的边界.     

乳腺肿块的超声声像图多元回归分析

目的：探讨超声声像图多元参数列乳腺良恶性肿块判别的意义。方法：回顾性分析382例经病理确诊乳腺良恶性肿块的超声声像图多元参数进行多元逐步回归分析。结果：诊断乳腺肿块的价值：内部回声〉钙化〉淋巴结〉后方声影〉边缘〉纵横比〉形态。回归方程：乳腺肿块良恶性情况=-1．40＋0．53（内部回声）＋0．49（淋巴结）＋0．16（形态）＋0．39（后方声影）＋0．25（边缘）＋0．21（纵横比）＋0．52（钙化），决定系数R^2=0．56。结论：综合乳腺超声声像图各参数。对判别乳腺肿块的良恶性有重要意义。     

Improving diffusion MRI using simultaneous multi-slice echo planar imaging

In diffusion MRI, simultaneous multi-slice single-shot EPI acquisitions have the potential to increase the number of diffusion directions obtained per unit time, allowing more diffusion encoding in high angular resolution diffusion imaging (HARDI) acquisitions. Nonetheless, unaliasing simultaneously acquired, closely spaced slices with parallel imaging methods can be difficult, leading to high g-factor penalties (i.e., lower SNR). The CAIPIRINHA technique was developed to reduce the g-factor in simultaneous multi-slice acquisitions by introducing inter-slice image shifts and thus increase the distance between aliased voxels. Because the CAIPIRINHA technique achieved this by controlling the phase of the RF excitations for each line of k-space, it is not directly applicable to single-shot EPI employed in conventional diffusion imaging. We adopt a recent gradient encoding method, which we termed "blipped-CAIPI", to create the image shifts needed to apply CAIPIRINHA to EPI. Here, we use pseudo-multiple replica SNR and bootstrapping metrics to assess the performance of the blipped-CAIPI method in 3× simultaneous multi-slice diffusion studies. Further, we introduce a novel image reconstruction method to reduce detrimental ghosting artifacts in these acquisitions. We show that data acquisition times for Q-ball and diffusion spectrum imaging (DSI) can be reduced 3-fold with a minor loss in SNR and with similar diffusion results compared to conventional acquisitions.     

基于各向异性扩散和小波变换的磁共振图像降噪和增强方法

降噪是医学图像处理中一个非常重要的问题,传统去噪方法在降低噪声的同时会模糊图像的边缘,各向异性扩散滤波在降低图像噪声的同时能够使图像的边缘得到保持.利用小波变换可以对图像进行多尺度分解,使我们可以在不同尺度上对图像进行处理.本文利用各向异性扩散滤波对MRI图像进行降噪,然后利用平稳小波变换对图像进行增强处理.实验结果表明,该方法在有效去除噪声的同时能够增强图像的细节,有效地提高了图像的质量.     

Knowledge-based tensor anisotropic diffusion of cardiac magnetic resonance images

We present a general formulation for a new knowledge-based approach to anisotropic diffusion of multi-valued and multi-dimensional images, with an illustrative application for the enhancement and segmentation of cardiac magnetic resonance (MR) images. In the proposed method all available information is incorporated through a new definition of the conductance function which differs from previous approaches in two aspects. First, we model the conductance as an explicit function of time and position, and not only of the differential structure of the image data. Inherent properties of the system (such as geometrical features or non-homogeneous data sampling) can therefore be taken into account by allowing the conductance function to vary depending on the location in the spatial and temporal coordinate space. Secondly, by defining the conductance as a second-rank tensor, the non-homogeneous diffusion equation gains a truly anisotropic character which is essential to emulate and handle certain aspects of complex data systems. The method presented is suitable for image enhancement and segmentation of single- or multi-valued images. We demonstrate the efficiency of the proposed framework by applying it to anatomical and velocity-encoded cine volumetric (4-D) MR images of the left ventricle. Spatial and temporal a priori knowledge about the shape and dynamics of the heart is incorporated into the diffusion process. We compare our results to those obtained with other diffusion schemes and exhibit the improvement in regions of the image with low contrast and low signal-to-noise ratio.     

Prospective and retrospective motion correction in diffusion magnetic resonance imaging of the human brain

Diffusion-weighting in magnetic resonance imaging (MRI) increases the sensitivity to molecular Brownian motion, providing insight in the micro-environment of the underlying tissue types and structures. At the same time, the diffusion weighting renders the scans sensitive to other motion, including bulk patient motion. Typically, several image volumes are needed to extract diffusion information, inducing also inter-volume motion susceptibility. Bulk motion is more likely during long acquisitions, as they appear in diffusion tensor, diffusion spectrum and q-ball imaging. Image registration methods are successfully used to correct for bulk motion in other MRI time series, but their performance in diffusion-weighted MRI is limited since diffusion weighting introduces strong signal and contrast changes between serial image volumes.In this work, we combine the capability of free induction decay (FID) navigators, providing information on object motion, with image registration methodology to prospectively - or optionally retrospectively - correct for motion in diffusion imaging of the human brain. Eight healthy subjects were instructed to perform small-scale voluntary head motion during clinical diffusion tensor imaging acquisitions.The implemented motion detection based on FID navigator signals is processed in real-time and provided an excellent detection performance of voluntary motion patterns even at a sub-millimetre scale (sensitivity ≥ 92%, specificity > 98%). Motion detection triggered an additional image volume acquisition with b = 0 s/mm² which was subsequently co-registered to a reference volume. In the prospective correction scenario, the calculated motion-parameters were applied to perform a real-time update of the gradient coordinate system to correct for the head movement.Quantitative analysis revealed that the motion correction implementation is capable to correct head motion in diffusion-weighted MRI to a level comparable to scans without voluntary head motion. The results indicate the potential of this method to improve image quality in diffusion-weighted MRI, a concept that can also be applied when highest diffusion weightings are performed.     

磁共振图像运动缺失修正算法研究

目的对在线扫描扩散成像法下由患者运动和磁场不均匀引起的MR图像运动缺失现象进行修正和补偿。方法首先检测运动缺失列,然后用相邻列对缺失列进行样条插值补偿,再提出改进算法,处理特殊的运动缺失情况。结果实验证明,该算法能对MR图像运动缺失现象进行修正和补偿,对背景噪声很强的缺失图像也有补偿效果。结论该算法在一定程度上提高了线扫描扩散成像法的图像质量。     

Atlas-based fiber reconstruction from diffusion tensor MRI data

Purpose

Develop a neural fiber reconstruction method based on diffusion tensor imaging, which is not as sensitive to user-defined regions of interest as streamline tractography.

Methods

A simulated annealing approach is employed to find a non-rigid transformation to map a fiber bundle from a fiber atlas to another fiber bundle, which minimizes a specific energy functional. The energy functional describes how well the transformed fiber bundle fits the patient??s diffusion tensor data.

Results

The feasibility of the method is demonstrated on a diffusion tensor software phantom. We analyze the behavior of the algorithm with respect to image noise and number of iterations. First results on the datasets of patients are presented.

Conclusions

The described method maps fiber bundles based on diffusion tensor data and shows high robustness to image noise. Future developments of the method should help simplify inter-subject comparisons of fiber bundles.     

Breast Ultrasound Despeckling Using Anisotropic Diffusion Guided by Texture Descriptors

Breast ultrasound (BUS) is considered the most important adjunct method to mammography for diagnosing cancer. However, this image modality suffers from an intrinsic artifact called speckle noise, which degrades spatial and contrast resolution and obscures the screened anatomy. Hence, it is necessary to reduce speckle artifacts before performing image analysis by means of computer-aided diagnosis systems, for example. In addition, the trade-off between smoothing level and preservation of lesion contour details should be addressed by speckle reduction schemes. In this scenario, we propose a BUS despeckling method based on anisotropic diffusion guided by Log–Gabor filters (ADLG). Because we assume that different breast tissues have distinct textures, in our approach we perform a multichannel decomposition of the BUS image using Log–Gabor filters. Next, the conduction coefficient of anisotropic diffusion filtering is computed using texture responses instead of intensity values as stated originally. The proposed algorithm is validated using both synthetic and real breast data sets, with 900 and 50 images, respectively. The performance measures are compared with four existing speckle reduction schemes based on anisotropic diffusion: conventional anisotropic diffusion filtering (CADF), speckle-reducing anisotropic diffusion (SRAD), texture-oriented anisotropic diffusion (TOAD), and interference-based speckle filtering followed by anisotropic diffusion (ISFAD). The validity metrics are the Pratt’s figure of merit, for synthetic images, and the mean radial distance (in pixels), for real sonographies. Figure of merit and mean radial distance indices should tend toward ‘1’ and ‘0’, respectively, to indicate adequate edge preservation. The results suggest that ADLG outperforms the four speckle removal filters compared with respect to simulated and real BUS images. For each method—ADLG, CADF, SRAD, TOAD and ISFAD—the figure of merit median values are 0.83, 0.40, 0.39, 0.51 and 0.59, and the mean radial distance median results are 4.19, 6.29, 6.39, 6.43 and 5.88.     

Regions of interest in FTIR imaging applications: diffusion of nicotine into ethylene-co-vinyl acetate films.

Effective image analysis of dynamic processes, such as diffusion and dissolution, requires precise reporting of component locations in space and time. An improved method for analyzing FTIR images is described which employs hypothesis testing in the spatial and temporal domains. Changes in the observed absorbance (over space and time) are revealed by comparison to a reference statistic, which can be tailored by choosing the size of a region of interest. This improved analysis method was used to compare the rates of diffusion of nicotine into poly(ethylene-co-vinyl acetate) film from aqueous solutions containing anionic and nonionic surfactants. Compared to a solution without surfactant, sodium dodecyl sulfate inhibited the uptake of nicotine from aqueous solution whereas Tween 40 enhanced the uptake. The nicotine diffusion rate also showed a dependence on the length of the hydrophobic segment of nonionic surfactants. These results demonstrate the roles of solubilization, wetting, and viscosity on diffusion-controlled drug release.     

Optical tractography of the mouse heart using polarization-sensitive optical coherence tomography

We developed a method to image myocardial fiber architecture in the mouse heart using a Jones matrix-based polarization-sensitive optical coherence tomography (PSOCT) system. The “cross-helical” laminar structure of myocardial fibers can be clearly visualized using this technology. The obtained myocardial fiber organization agrees well with existing knowledge acquired using conventional histology and diffusion tensor magnetic resonance imaging.OCIS codes: (110.4500) Optical coherence tomography, (230.5440) Polarization     

Texture preservation and speckle reduction in poor optical coherence tomography using the convolutional neural network

For a poor quality optical coherence tomography (OCT) image, quality enhancement is limited to speckle residue and edge blur as well as texture loss, especially at the background region near edges. To solve this problem, in this paper we propose a de-speckling method based on the convolutional neural network (CNN). In the proposed method, we use a deep nonlinear CNN mapping model in the serial architecture, here named as OCTNet. Our OCTNet in the proposed method can fully utilize the deep information on speckles and edges as well as fine textures of an original OCT image. And also we construct an available pertinent dataset by combining three existing methods to train the model. With the proposed method, we can accurately get the speckle noise from an original OCT image. We test our method on four experimental human retinal OCT images and also compare it with three state-of-the-art methods, including the adaptive complex diffusion (ACD) method and the curvelet shrinkage (Curvelet) method as well as the shearlet-based total variation (STV) method. The performance of these methods is quantitatively evaluated in terms of image distinguishability, contrast, smoothness and edge sharpness, and also qualitatively analyzed at aspects of speckle reduction, texture protection and edge preservation. The experimental results show that our OCTNet can reduce the speckle noise and protect the structural information as well as preserve the edge features effectively and simultaneously, even where the background region near edges. And also our OCTNet has full advantages on excellent generalization, adaptiveness, robust and batch performance. These advantages make our method be suitable to process a great mass of different images rapidly without any parameter fine-turning under a time-constrained real-time situation.