基于图像内容层次表征的遥感图像 分割方法

提出基于图像内容层次表征的高分辨率遥感图像快速多精度分割方法。首先根据初始分割结果建立区域邻接图(RAG),并将其定义为马尔可夫随机场(MRF);然后引入光谱、形状和边缘等图像特征进行层次合并,通过记录层次合并过程获得图像内容的层次表征;最后根据层次表征中不同层级对象之间的关系快速生成任意不同精度的分割结果,以满足不同应用的需求。利用QuickBird卫星图像进行实验和评价的结果表明,本文方法具有较高的精度和效率。     

均衡化特征匹配的非刚体细胞形态跟踪

提出利用均衡化特征匹配来进行非刚性细胞形体跟踪的方法。采用重启动的随机游走方法建立并求解特征匹配概率模型,利用双向均衡方法对匹配邻接矩阵进行均衡化处理,得到指定目标与待跟踪目标之间的精确匹配,以获得目标的定位跟踪结果。同时利用特征匹配结果进行目标的自动标定,并应用图像分割方法进行目标的精确轮廓跟踪。实验结果表明,将该方法应用于视频中动态背景下的运动细胞形态跟踪时,在背景相似度较高及目标迅速移动的条件下,表现出了良好的性能,与同类方法相比可获得较高的定位精度以及更为准确的目标轮廓。     

改进FCM在交互式图像分割中的应用

为解决由于自然纹理的干扰而导致的分割图像边缘模糊问题,对模糊C均值聚类算法进行改进并应用于交互式图像分割中。用户通过输入种子点来获得目标和背景的主要特征,并将输入的种子点作为聚类中心点;提出全局空间相似性度量标准并引入Gabor能量滤波器来计算图像中各点到聚类中心的距离;算法首次引入边缘密度概念定义权重因子,根据图像特点,自适应地计算图像中任意一点的纹理特征和颜色特征在特征空间中所占比例,使得到的特征更加准确地描述图像的本质属性。对具有自然纹理背景的图像进行仿真实验,应用两种性能指标来比较本文所提算法与随机游走算法的分割精度。实验结果表明,本文算法分割精度高于模糊聚类和随机游走算法。     

医学图像中粘连细胞分割方法研究

研究粘连细胞图像的分割问题。针对医学细胞图像中,细胞经常发生粘连,细胞粘连部分由于发生像素灰度混合,造成图像不清晰。传统的应用灰度的分割算法由于不能很好的区分粘连部分的灰度,不能完整的分割细胞的问题。为了解决上述问题,提出了自适应阀值调整的粘连细胞分割方法。通过求得粘连部位的细胞像素的具体特征,运用阀值的自适应变化区分粘连部位像素之间的不同,对像素进行划分,克服了传统方法像素划分不清的弊端。实验表明,方法能够有效的分割医学图像中的粘连细胞,取得了比较好的效果。     

一种新的红外图像快速分割方法研究

研究红外图像目标分割快速优化问题。在高分辨率红外图像的分割中,红外图像存在数据量大、目标的边缘模糊和噪声较大等导致分辨率低和实时性差。为了快速准确分割,提出基于低尺度分割阈值预测的快速红外图像分割方法(LFIRS)。首先建立尺度阶数计算模型以确定保留原始红外图像目标基本信息所需的最小尺度,在分析多尺度过程对具有目标/背景强相关特性的红外图像进行分割阈值,建立多尺度红外图像分割阈值的相关模型(CLM),结合经典二维阈值法得到的最低两个尺度的阈值CLM模型参数,可以通过CLM模型与最小尺度的快速获取原始尺度的分割阈值,实现红外图像的快速分割。实验结果表明,改进方法提高了图像分辨和分割速度,且改善了分割效果。     

基于视觉的多特征手势识别

手势是一种自然直观的交互方式,基于视觉的手势识别是实现新一代人机交互的关键技术。本文在已有的手势识别技术基础上,从手势分割及手势表示两方面着手,提出了一种单目视觉下的手势识别方法。利用颜色特征检测肤色区域,成功分割出人手;利用人手的轮廓及凸缺陷检测指尖,再利用指尖的数目和方位来表示一个手势,进而结合轮廓长度和面积等几何特征完成手势识别。传统的指尖检测方法需要遍历并扫描手掌外轮廓,计算量大,本文通过凸缺陷检测指尖,减少了计算量,提高了指尖检测的速度。实验结果表明,本文的方法具有很好的鲁棒性及实时性,能适应环境的变化。     

基于卡尔曼滤波改进的精子图像序列分割方法

图像分割是精子图像识别的一项关键技术,在精子运动能力分析中起着至关重要的作用。本文对采集的连续精子图像序列进行灰度化、去噪等预处理后,采用Otsu算法对首幅动物精子图像二值化,对后续图像采用Kalman Filter确定二值化阈值范围,改进Otsu算法求出每一幅图像的适当阈值并进行二值化,缩短算法时间并能保证分割精度。应用形态学消除精子尾部和部分精子之间的粘连现象,通过计算和比较目标面积、形状因子,去除小颗粒杂质以及形状及灰度和精子相似的杂质,为精子运动能力检测提供高质量的分割图像。     

mRNA and miRNA Expression Analyses of the MYC/E2F/miR-17-92 Network in the Most Common Pediatric Brain Tumors

Numerous molecular factors disrupt the correctness of the cell cycle process leading to the development of cancer due to increased cell proliferation. Among known causative factors of such process is abnormal gene expression. Nowadays in the light of current knowledge such alterations are frequently considered in the context of mRNA–miRNA correlation. One of the molecular factors with potential value in tumorigenesis is the feedback loop between MYC and E2F genes in which miR-17-5p and miR-20a from the miR-17-92 cluster are involved. The current literature shows that overexpression of the members of the OncomiR-1 are involved in the development of many solid tumors. In the present work, we investigated the expression of components of the MYC/E2F/miR-17-92 network and their closely related elements including members of MYC and E2F families and miRNAs from two paralogs of miR-17-92: miR-106b-25 and miR-106a-363, in the most common brain tumors of childhood, pilocytic astrocytoma (PA), WHO grade 1; ependymoma (EP), WHO grade 2; and medulloblastoma (MB), WHO grade 4. We showed that the highest gene expression was observed in the MYC family for MYCN and in the E2F family for E2F2. Positive correlation was observed between the gene expression and tumor grade and type, with the highest expression being noted for medulloblastomas, followed by ependymomas, and the lowest for pilocytic astrocytomas. Most members of miR-17-92, miR-106a-363 and miR-106b-25 clusters were upregulated and the highest expression was noted for miR-18a and miR-18b. The rest of the miRNAs, including miR-19a, miR-92a, miR-106a, miR-93, or miR-25 also showed high values. miR-17-5p, miR-20a obtained a high level of expression in medulloblastomas and ependymomas, while close to the control in the pilocytic astrocytoma samples. miRNA expression also depended on tumor grade and histology.     

The Effect of Hypoxia on the Expression of CXC Chemokines and CXC Chemokine Receptors—A Review of Literature

Hypoxia is an integral component of the tumor microenvironment. Either as chronic or cycling hypoxia, it exerts a similar effect on cancer processes by activating hypoxia-inducible factor-1 (HIF-1) and nuclear factor (NF-κB), with cycling hypoxia showing a stronger proinflammatory influence. One of the systems affected by hypoxia is the CXC chemokine system. This paper reviews all available information on hypoxia-induced changes in the expression of all CXC chemokines (CXCL1, CXCL2, CXCL3, CXCL4, CXCL5, CXCL6, CXCL7, CXCL8 (IL-8), CXCL9, CXCL10, CXCL11, CXCL12 (SDF-1), CXCL13, CXCL14, CXCL15, CXCL16, CXCL17) as well as CXC chemokine receptors—CXCR1, CXCR2, CXCR3, CXCR4, CXCR5, CXCR6, CXCR7 and CXCR8. First, we present basic information on the effect of these chemoattractant cytokines on cancer processes. We then discuss the effect of hypoxia-induced changes on CXC chemokine expression on the angiogenesis, lymphangiogenesis and recruitment of various cells to the tumor niche, including myeloid-derived suppressor cells (MDSCs), tumor-associated macrophages (TAMs), tumor-associated neutrophils (TANs), regulatory T cells (T_regs) and tumor-infiltrating lymphocytes (TILs). Finally, the review summarizes data on the use of drugs targeting the CXC chemokine system in cancer therapies.     

The Role of Tumor Microenvironment Cells in Colorectal Cancer (CRC) Cachexia

Cancer cachexia (CC) is a multifactorial syndrome in patients with advanced cancer characterized by weight loss via skeletal-muscle and adipose-tissue atrophy, catabolic activity, and systemic inflammation. CC is correlated with functional impairment, reduced therapeutic responsiveness, and poor prognosis, and is a major cause of death in cancer patients. In colorectal cancer (CRC), cachexia affects around 50–61% of patients, but remains overlooked, understudied, and uncured. The mechanisms driving CC are not fully understood but are related, at least in part, to the local and systemic immune response to the tumor. Accumulating evidence demonstrates a significant role of tumor microenvironment (TME) cells (e.g., macrophages, neutrophils, and fibroblasts) in both cancer progression and tumor-induced cachexia, through the production of multiple procachectic factors. The most important role in CRC-associated cachexia is played by pro-inflammatory cytokines, including the tumor necrosis factor α (TNFα), originally known as cachectin, Interleukin (IL)-1, IL-6, and certain chemokines (e.g., IL-8). Heterogeneous CRC cells themselves also produce numerous cytokines (including chemokines), as well as novel factors called “cachexokines”. The tumor microenvironment (TME) contributes to systemic inflammation and increased oxidative stress and fibrosis. This review summarizes the current knowledge on the role of TME cellular components in CRC-associated cachexia, as well as discusses the potential role of selected mediators secreted by colorectal cancer cells in cooperation with tumor-associated immune and non-immune cells of tumor microenvironment in inducing or potentiating cancer cachexia. This knowledge serves to aid the understanding of the mechanisms of this process, as well as prevent its consequences.