乳腺肿瘤超声图像中感兴趣区域的自动检测

由于斑点噪声、伪影以及病灶形状多变的影响,乳腺肿瘤超声图像中肿瘤区域的自动检测以及病灶的边缘提取比较困难,已有的方法主要是由医生先手工提取感兴趣区域(ROI)。本研究提出一种乳腺肿瘤超声图像中感兴趣区域自动检测的方法,选用超声图像的局部纹理、局部灰度共生矩阵以及位置信息作为特征,采用自组织映射神经网络进行分类,自动识别乳腺肿瘤区域。对包含168幅乳腺肿瘤超声图像的数据库进行识别的结果表明:该方法自动识别ROI的准确率达到86.9%,可辅助医生提取肿瘤的实际边缘以及进一步诊断。     

乳腺钼靶在女职工乳腺普查中的X线表现与应用价值分析

目的分析乳腺钼靶在女职工乳腺普查中的X线表现和应用价值。方法资料选自2013年9月~12月在本院行乳腺普查的1000例已婚妇女职工为研究对象,分析本组受检者钼靶X线和应用价值。结果不同年龄段受检者乳腺普查即乳腺增生、乳腺纤维腺瘤、单纯性囊肿、乳腺癌检出率均不同,其中年龄在36~55岁内乳腺癌发生率较高;经手术病理结果证实确诊率81.82%,误诊率18.18%,漏诊率0.00%。结论乳腺钼靶在女职工乳腺疾病普查中的X线表现良好,可清晰显示出乳腺疾病X线片且符合率较高,具有积极临床应用和研究价值。     

基于小波变换的乳腺X线影像微钙化点感兴趣区域提取新技术

为乳腺癌早期诊断和乳腺X线影像微钙化点计算机辅助检测作前期预处理,提出了一种基于小波变换的微钙化点感兴趣区提取新技术。其具体思路是:(1)将乳腺区域图像提取成等大的子图像;(2 )对每一幅子图像进行小波变换,根据特征参数ρ讨论最优小波变化参数和阈值T;(3)根据阈值T判别子图像是否属于感兴趣区。对临床实际病例(2 0幅乳腺X线影像)的试验结果表明,该方法具有较高的检出率(89.7% ) ,和较为满意的假阳性率(2 .1% )。     

基于乳腺钼靶侧斜位图像的近期乳腺癌风险预测研究

探讨乳腺钼靶侧斜位图像双侧不对称性与近期乳腺癌风险之间的关联性。回顾性分析556例连续两年进行乳腺普查的样本,当前年患癌与阴性样本各278例,且两组样本年龄匹配。对前一年乳腺钼靶侧斜位图像进行预处理获取双侧局部对应区域及全局区域后分别提取空间差异性、结构相似性及位置信息等85维影像特征。删除相关性较高的特征后剩余78个特征,采用留一法与逐步回归分析进行特征选择,并建立基于广义线性模型的近期乳腺癌风险预测模型。结果显示,全局及局部不对称性特征相结合进行风险预测AUC值为0.666 7±0.022 6,对其进行混淆矩阵分析,特异性为0.690 6,灵敏度为0.521 6。对风险预测值进行回归分析,优势比(Odds Ratio)随风险的增加而显著增加,P值为0.002 033。3个基于年龄的子集(37~49岁,50~65岁,66~87岁)进行近期乳腺癌风险预测,AUC值分别为0.681 0±0.043 2,0.671 6±0.030 0及0.678 2±0.054 7,对应3个年龄段的特异性分别为0.702 7,0.694 3和0.723 4,灵敏度分别为0.554 1,0.490 4和0.574 5。两个基于乳腺X线分型的子集(BIRADS 2、BIRADS 3) AUC值分别为0.654 5±0.036 9及0.694 4±0.03,特异性分别为0.676 2和0.733 3,灵敏度分别为0.522 9和0.536 9。结果表明,乳腺钼靶侧斜位图像全局及局部不对称性相结合对近期乳腺癌风险预测具有潜在有效价值。     

基于深度学习的乳腺钼靶图像分类方法研究进展

钼靶检查是当前临床诊断乳腺肿瘤的常规手段,患者痛苦相对较小、简便易行、分辨率高、可重复性好。为了提高诊断效率,减小误诊风险,针对乳腺钼靶图像开发基于人工智能的计算机辅助诊断系统(computer-aided diagnosis,CAD)显得尤为重要。传统的分类方法需要使用大量的手工特征,而深度学习能够自动从数据中学习特征,避免了传统算法中人工设计、提取特征的复杂性和局限性。我们从感兴趣区域和全图两个方面对近年来基于深度学习的乳腺钼靶图像分类方法研究进展予以综述和展望。调研发现深度学习在乳腺钼靶图像分类方面展示了不错的效果,其中基于深度卷积神经网络的分类方法已经成为当下的热门技术。     

乳腺钼靶X线影像中结构扭曲检测技术研究的进展

乳腺癌是全球女性发病率最高的恶性肿瘤,通过筛查实现早期发现、早期诊断和早期治疗对降低乳腺癌死亡率至关重要。乳腺钼靶X线摄影术是目前最普遍适用的乳腺癌筛查方法。有效检测乳腺钼靶X线影像中的结构扭曲病灶有利于提高筛查的质量和效率。就目前乳腺钼靶X线影像中结构扭曲检测技术的研究现状、存在的问题和发展趋势进行了综述。     

基于双视角和多分类器信息融合的乳腺钼靶图像肿块分类研究

乳腺肿块良恶性分类是计算机辅助诊断(CAD)的重要环节,如何提高分类的正确率和稳定性是分类研究的重点.本研究提出了4种基于双视角和多分类器信息融合的乳腺钼靶图像肿块分类模式.其中,模式1是单视角下的多分类器融合;模式2是分别先对每个分类器在两个视角下的输出进行视角融合,再对其融合结果进行多分类器融合;模式3是分别先在每...     

乳腺钼靶X线影像中结构扭曲的特征提取研究

本研究以灰度共生矩阵描述乳腺钼靶X线影像中结构扭曲的纹理特征.对学习样本（乳腺结构扭曲样本44个,正常样本78个）,计算五个反映纹理性质的特征参数,根据相应的Fisher系数,确定最适合作为分类依据的特征参数或特征参数组合.用线性判别分析对测试样本（乳腺结构扭曲样本43个,正常样本78个）进行分类.分类结果表明本研究确定的纹理特征熵（ENT）是识别乳腺结构扭曲的最佳统计参数（分类正确率达78.5%、ROC曲线下的面积为0.786）.     

采用多尺度多级组合分类器快速定位乳腺X片中的感兴趣区域

乳腺癌是妇女常见恶性肿瘤之一,早期诊断和治疗是降低乳腺癌患者死亡率的关键. 微钙化是乳腺癌早期的一个重要标志,因此快速准确地找出乳腺X光片中含有微钙化簇的感兴趣区域(ROI)是成功诊断的第一步.乳腺X光片中含有大量无病变区域和少量微钙化区域,形成了一种典型的不对称分类问题.本研究结合大量无病变区域的信息训练多级组合分类器,并借助多尺度方法加快筛选速度,以定位ROI.在真实的数字化X线乳腺照片上的实验表明,该方法在无漏检的情况下,可以排除92.64%的正常区域,而且基于Matlab处理,对于每幅图片的平均处理时间仅为7 s.     

乳腺钼靶片中钙化灶对乳腺疾病的诊断价值分析

目的分析乳腺钼靶片中钙化灶对乳腺疾病的诊断价值。方法选取我院2011年3月~2014年3月收治的经手术或穿刺活检病理证实的58例乳腺疾病病灶钙化患者,所有患者均行乳腺钼靶片检查,分析乳腺良恶性疾病的钙化特点。结果本组58例病灶中,良性病变23例,占39.7%,其中乳腺纤维瘤14例,乳腺囊肿5例,乳腺血管粥样硬化2例,乳腺增生2例,恶性病变35例,占60.3%,其中浸润导管癌18例,腺癌13例,黏液癌4例。良性钙化可见大颗粒钙化,爆米花样钙化、弧形钙化、圆形钙化,恶性钙化可见杆状、分支状及泥沙样多形性、泥沙样钙化、小杆状及分支样。单位面积钙化数目≥5枚/cm252例,其中47例为恶性病变,单位面积钙化数目<5枚/cm26例,均为良性病变。结论乳腺钼靶片中钙化灶对乳腺疾病的诊断以及乳腺疾病良恶性的鉴别具有重要的意义,能够降低误诊和漏诊的发生率,为临床治疗提供重要依据。     

基于轮廓匹配和多项式拟合的虹膜分割算法

虹膜图像在采集过程中由于个体差异、采集环境等因素的影响,存在遮挡、偏角等噪声。针对该问题,提出一种基于轮廓匹配和多项式拟合的虹膜分割方法。首先将虹膜图像分为理想型虹膜图像与非理想型虹膜图像;其次针对理想型虹膜图像,根据圆的标准方程检测虹膜边界实现虹膜分割,针对非理想型虹膜图像,通过轮廓匹配算法获得带遮挡部分的虹膜后,采用多项式拟合的方法去除眼睑等遮挡部分,实现虹膜的准确分割。本文算法实现虹膜分割的错误率nice1为1.5%,nice2为1.9%,且F1值达到93.65%。结果表明,该方法能够有效地去除虹膜图像中的遮挡噪声,具有较高的虹膜分割准确率。     

Correlation between signal-to-noise ratios and region of interest sensitivity ratios of bipolar EEG measurements

We have developed a parameter, which describes how well the measurement is concentrated on the region of interest source area compared to other source areas in the volume conductor. The parameter concept is called the region of interest sensitivity ratio (ROISR). We assume that ROISR is also connected to the SNR of the measurement. The objective of the present study was to investigate the assumed correlation between the ROISR and SNR of the measurement with three-layer spherical head model. We studied how the source distribution and orientation affect the correlation and thus how applicable the ROISR is in analysing the sensitivity distributions of measurements. We simulated bipolar EEG-evoked potential measurements with 16 combinations of four-source distribution and four-source orientation models. The results indicate that the ROISR correlates with the SNR of the measurement with all tested source distributions and orientations. Thus the ROISRs concept can be applied to analyse measurement setups by modelling and analysing the sensitivity distributions.     

Automatic segmentation algorithm for the extraction of lumen region and boundary from endoscopic images

A new segmentation algorithm for lumen region detection and boundary extraction from gastro-intestinal (GI) images is presented. The proposed algorithm consists of two steps. First, a preliminary region of interest (ROI) representing the GI lumen is segmented by an adaptive progressive thresholding (APT) technique. Then, an adaptive filter, the Iris filter, is applied to the ROI to determine the actual region. It has been observed that the combined APT-Iris filter technique can enhance and detect the unclear boundaries in the lumen region of GI images and thus produces a more accurate lumen region, compared with the existing techniques. Experiments are carried out to determine the maximum error on the extracted boundary with respect to an expert-annotated boundary technique. Investigations show that, based on the experimental results obtained from 50 endoscopic images, the maximum error is reduced by up to 72 pixels for a 256 × 256 image representation compared with other existing techniques. In addition, a new boundary extraction algorithm, based on a heuristic search on the neighbourhood pixels, is employed to obtain a connected single pixel width outer boundary using two preferential sequence windows. Experimental results are also presented to justify the effectiveness of the proposed algorithm.     

脊柱结核和恶性肿瘤弥散加权成像中感兴趣区设置方法初探

为了探讨脊柱结核和脊柱恶性肿瘤MR弥散加权成像(DWI)中感兴趣区(ROI)设置对表观弥散系数(ADC)值测量的影响,为ADC值测量选择较为合理、有效的测量方法。我们收集2009年8月-2010年6月间32例经病理和临床随访证实的脊柱结核和恶性肿瘤患者的常规MRI和DWI资料。以常规MRI为指导,在DWI图像中病变最大层面,分别以整个病变区域取单个ROI和在整个病变区域内取面积相同的多个ROI,测量其信号强度(SI)和ADC值并进行统计学分析。由受试者工作特性(ROC)曲线确定临界点后,比较两种方法所测得的ADC值诊断脊柱结核的灵敏度、特异度和准确率。结果发现单个ROI测得的ADC均值较多个ROI低,但差异没有统计学意义(P>0.05)。单个ROI或多个ROI时脊柱结核与肿瘤的ADC均值差别都有统计学意义(P<0.05),多个ROI测得的ADC值诊断脊柱结核的灵敏度、特异度和准确率高于前者。因此,对于脊柱结核和肿瘤,多个ROI测得的ADC值较单个ROI测得的ADC值能更好地反映水分子的弥散状态,在脊柱结核与恶性肿瘤的鉴别诊断中更有价值。     

核磁共振图像分割算法及应用进展

核磁共振成像(MRI)以其无辐射、多方位成像、空间分辨率高等优点在影像医学领域广泛应用,核磁共振图像的分割也发挥着越来越重要的作用.对应用较广的核磁图像的分割算法的原理和应用进行了系统的综述,将核磁图像分割算法分为5个主要研究方面:基于阈值,基于模式识别,基于活动轮廓模型,基于马尔科夫随机场(MRF),基于图切割;给出了不同算法分割特点和相关应用范围,并将部分算法应用在腹部核磁图像上进行分割实验,展示了不同算法分割核磁图像后的效果和特点.最后,展望了核磁图像分割的未来的发展趋势.     

Melanomas non-invasive diagnosis application based on the ABCD rule and pattern recognition image processing algorithms

In this paper an automated dermatological tool for the parameterization of melanomas is presented. The system is based on the standard ABCD Rule and dermatological Pattern Recognition protocols. On the one hand, a complete stack of algorithms for the asymmetry, border, color, and diameter parameterization were developed. On the other hand, three automatic algorithms for digital image processing have been developed in order to detect the appropriate patterns. These allow one to calculate certain quantitative features based on the aspect and inner patterns of the melanoma using simple-operation algorithms, in order to minimize response time. The database used consists of 160 500×500-pixel RGB images (20 images per pattern) cataloged by dermatologists, and the results have turned out to be successful according to assessment by medical experts. While the ABCD algorithms are mathematically reliable, the proposed algorithms for pattern recognition produced a remarkable rate of globular, reticular, and blue veil Pattern recognition, with an average above 85% of accuracy. It thus proves to be a reliable system when performing a diagnosis.     

加权SVM在中医舌象分类与识别中的应用研究

加权SVM是解决不均衡舌象样本分类和识别的有效方法，但在边界支撑向量很少是错分样本的情况下，已有方法有时不能取得良好的效果。本研究首先采用已有的加权方法进行舌象样本的分类与识别，在训练结果不能使得大小样本的分类率得到很好均衡的情形下，分别进一步加大和缩小重要样本和非重要样本的惩罚项，直到满足要求为止。理论分析和实验均表明该方法的有效性。     

An investigation of matching symmetry in the human pinnae with possible implications for 3D ear recognition and sound localization

The human external ears, or pinnae, have an intriguing shape and, like most parts of the human external body, bilateral symmetry is observed between left and right. It is a well-known part of our auditory sensory system and mediates the spatial localization of incoming sounds in 3D from monaural cues due to its shape-specific filtering as well as binaural cues due to the paired bilateral locations of the left and right ears. Another less broadly appreciated aspect of the human pinna shape is its uniqueness from one individual to another, which is on the level of what is seen in fingerprints and facial features. This makes pinnae very useful in human identification, which is of great interest in biometrics and forensics. Anatomically, the type of symmetry observed is known as matching symmetry, with structures present as separate mirror copies on both sides of the body, and in this work we report the first such investigation of the human pinna in 3D. Within the framework of geometric morphometrics, we started by partitioning ear shape, represented in a spatially dense way, into patterns of symmetry and asymmetry, following a two-factor anova design. Matching symmetry was measured in all substructures of the pinna anatomy. However, substructures that ‘stick out’ such as the helix, tragus, and lobule also contained a fair degree of asymmetry. In contrast, substructures such as the conchae, antitragus, and antihelix expressed relatively stronger degrees of symmetric variation in relation to their levels of asymmetry. Insights gained from this study were injected into an accompanying identification setup exploiting matching symmetry where improved performance is demonstrated. Finally, possible implications of the results in the context of ear recognition as well as sound localization are discussed.     

Evaluation of the class II region of the major histocompatibility complex of the greyhound with the genomic matching technique and sequence-based typing

The class II region of the major histocompatibility complex was evaluated in 25 greyhounds by sequence-based typing and the genomic matching technique (GMT). Two new DLA-DRB1 alleles were identified. Twenty-four dogs carried the DLA-DRB1*01201/DQA1*00401/DQB1*01303/DQB1*01701 haplotype, which carries two DQB1 alleles. One haplotype was identified from which DQB1 and DQA1 appeared to be deleted. The GMT enabled detection of DQB1 copy number, discrimination of the different class II haplotypes and the identification of new, possibly biologically relevant polymorphisms.     

The effect of target modality on visual and proprioceptive contributions to the control of movement distance

The goal of this study was to determine whether the sensory nature of a target influences the roles of vision and proprioception in the planning of movement distance. Two groups of subjects made rapid, elbow extension movements, either toward a visual target or toward the index fingertip of the unseen opposite hand. Visual feedback of the reaching index fingertip was only available before movement onset. Using a virtual reality display, we randomly introduced a discrepancy between actual and virtual (cursor) fingertip location. When subjects reached toward the visual target, movement distance varied with changes in visual information about initial hand position. For the proprioceptive target, movement distance varied mostly with changes in proprioceptive information about initial position. The effect of target modality was already present at the time of peak acceleration, indicating that this effect include feedforward processes. Our results suggest that the relative contributions of vision and proprioception to motor planning can change, depending on the modality in which task relevant information is represented.