The increasing availability of protein-protein interaction data makes network alignment more and more important in predicting new functions of proteins and inferring the evolutionary history of protein interaction networks.However,most present methods either ignore the node or structure information,or adopt heuristics.The authors present an exact network alignment algorithm by transforming network comparison into a linear programming problem.A powerful mathematical programming optimizer ILOG CPLEX is used to solve the linear programming problem.

Combining the Brownian ratchet models and the power stroke models,the authors proposed a new mathematical scheme to quantitatively describe the dynamics of molecular motors.That was,a set of functions,periodic either in time or in space,was introduced to modeling the power strokes.Analytic solutions for the probability current are achieved and the relationships of the current with load and asymmetry were seen about.The results were consistent with experimental data.There was no need for the interaction potential between the molecular motor and its track was asymmetric in this model,which makes it more robust than the Brownian ratchet models.

A new algorithm of image layer-presentation was proposed. The key concept of the algorithm was in that the image grayscale function f(x,y), which was comparatively irregular, was approximated by a series of high-regular grayscale functions gn(x,y). The algorithm had the feature of fast convergence and so it was a good approximator. Due to its high-regularity, gn might be stored in a quite small memory space. Thus, the algorithm gives a new way for data processing and image reconstruction.

A method based on the model based neural network (MBNN) is proposed to automatically analyze high-resolution G bands of Triticum monococcum chromosomes. The MBNN-3P is employed to segmented the G-band images. Then five features of chromosomes are extracted from the segmented images. At last, the features are input into the MBNN for classification. The results indicate that the method provides a significant way for the automatic analysis of high-resolution G bands of plant chromosomes accurately.

Recognition of lung cancer cells is very important to the clinical diagnosis of lung cancer. In this paper we present a novel method to extract the structure characteristics of lung cancer cells and automatically recognize their types. Firstly soft mathematical morphology methods are used to enhance the grayscale image, to improve the definition of images, and to eliminate most of disturbance, noise and information of subordinate images, so the contour of target lung cancer cell and biological shape characteristic parameters can be extracted accurately. Then the minimum distance classifier is introduced to realize the automatic recognition of different types of lung cancer cells. A software system named "CANCER.LUNG" is established to demonstrate the efficiency of this method. The clinical experiments show that this method can accurately and objectively recognize the type of lung cancer cells, which can significantly improve the pathology research on the pathological changes of lung cancer and clinical assistant diagnoses.

This paper introduces a three-dimensional (3D) reconstruction algorithm of the brain stem nuclei based on fast centroid auto-registration. The research is based on methods and theories of computer stereo vision, and by image information processing three-point pattern local search, registration and auto-tracing for the centroids of the brain stem nuclei were accomplished. We adopt two-peak threshold, edge detection and grayscale image enhancement to extract contours of the nuclei's structures. The experimental results obtain the spatial structure information and 3D image of the brain stem nuclei, show spatial relationship between 14 pairs of nuclei, and quantitate morphological parameters of each type of nuclei's 3D structure. This work is significant to neuroanatomy research and clinic applications. Furthermore, a software system named BRAIN.HUK is established.

鉴于非刚体的运动分析业已成为计算机视觉中的一个重要应用领域,为了使人们对该领域的研究现状有个概略了解,首先基于微分几何上的高斯曲率变化,对不同非刚性物体(简称非刚体)进行了分类,指出可把所有运动物体分为8类;然后对该领域目前存在的各种算法进行归纳总结,指出可把它们分为基于特征的方法和基于形状模型的方法两大类,并讨论了这两类方法各自的优势和不足之处;最后分析了该领域面临的困难,并展望了它未来可能的发展方向.同时指出,非刚性运动的视觉分析虽是一个蓬勃发展的研究领域,但目前仍处于初期阶段,因为近年来所进行的工作只是涉及众多困难问题中的较少部分,而且现有的各种模型和算法都还很不完善,人们还远未找到解决非刚性运动视觉分析问题的最有效的途径,但已有成果表明,一些相关研究领域,如语音识别和计算机图形学将会对该领域的发展提供帮助.

Contextual information and a priori knowledge play important roles in image segmentation based on neural networks. This paper proposed a method for including contextual information in a model-based neural network (MBNN) that has the advantage of combining a priori knowledge. This is achieved by including Markov random field (MRF) into the MBNN and this novel neural network is termed as MRF-MBNN. Then the proposed method is applied to segmenting the images. Experimental results indicate the MRF-MBNN is superior to the MBNN in image segmentation. This study is a successful attempt of incorporating contextual information and a prior knowledge into neural networks to segment images.

提出了一种改进的基于模型的神经网络(MBNN,model-based neural network)图像分割算法.采用马尔科夫随机场(MRF,Markov randomfield)对图像建模,将该模型融入MBNN,应用改进的最大期望值(EM,expectationmaximization)算法估计网络中MRF参数,采用预指定类别数技术减少网络计算量,最终实现图像分割.实验结果表明,该算法能够有机地将先验知识和图像局部统计相关性结合起来,从而有效地完成图像分割.

Biomolecular motors are tiny engines that transport materials at the microscopic level within biological cells. In recent years, Elston and Peskin et al have investigated the effect of the elastic properties of the tether that connects the motor to its cargo at the speed of the motor. In this paper we extend their work and present a tether in the form of symmetric linear potential. Our results show that when the driving mechanism is an imperfect Brownian ratchet, the average speed decreases as the stiffness of the tether increases in the limit of large motor diffusion coefficient, which is similar to the results of Elston and Peskin. However, a threshold for the stiffness of the tether connecting the motor to its cargo is found in our model. Only when the tether is stiffer than the threshold can the motor and its cargo function co-operatively, otherwise, the motor and its cargo depart from each other. This result is more realistic than that of the spring model of Elston and Peskin.

生物医学图像在成像时不可避免地受到噪声影响,因此噪声去除是生物医学图像处理的一项重要研究课题.将小波神经网络引入图像去噪领域中,通过多种技术优化网络学习过程,最终建立一种图像去噪新算法.实验结果表明,该算法在去除噪声上优于传统的中值滤波等方法,并具有较强的鲁棒性;同时能够最大限度地保护图像的细节信息,具有很好的保真度.

In this paper the recognition of Small Cell Carcinoma (SCC) is studied. For each type we select 128 samples for training, and randomly measure 200 cells in each sample. We introduce multi-scale morphology based on centroid coordinates to extract the boundaries of nuclei and obtain feature images of nuclei. The features of lung cancer cells are described by morphological and colorimetrical parameters, which is valuable to recognize SCC. Then the architecture of self-organizing feature mapping (SOFM) neural network is studied for recognition of SCC. The weights of the network are adjusted by self-organizing competition, and finally inputted patterns are classified. This algorithm has the advantage of parallelism and fast-convergence, and may simplify the analysis of SCC. Clinical experiment results show that the correctness ratio of this system may reach 95.3% while recognizing lung cancer cell types. Our work is significant to the pathological researches of lung cancer, assistant clinic diagnosis, and assessment of therapeutic effects. Meanwhile a software system named as SCC. LUNG is established for automatic analysis.

Ultrastructure of nucleolar DNA is most important to elucidate transcription site of rRNA. Many judgments were made in this respect but all of them were based on direct experimental observations. A new technology for the quantitative analysis of the location of nucleolar DNA in situ from EM data of the ultrastructure of the nucleolus in allium sativum cells and from the modified NAMA-Ur DNA specific staining, method was developed. For this purpose, a new methodology of multi-scale morphological gradient operator and automatic analysis software named "HEREN. CELL" were created and adopted. By using this method, superfine contour of the nucleolar ultrastruture was obtained and showed the location and location of nucleolar DNA in situ. From the results, a "petaline envelope" model was proposed, which indicated that nucleolar DNA fibrils radiate from center to periphery of the nucleolus. The study will contribute to the development of the EM image processing technology and the study of nucleolar ultrastructure of the nucleolus.

Successful segmentation of chromosomes is important for their analysis. Isolation of chromosomes is the major topic of the segmentation in previous studies, but it is at low resolution. While the model-based neural network (MBNN), in this paper, is adopted to segment the G bands of Triticum monococcum chromosomes, which is a high-resolution approach. Combined with the MBNN as a core technique, various auxiliary techniques have been used to find out the optimal method. A series of experimental results indicate that the optimal method for segmentation of G bands, here termed as MBNN-3P, is that using the MBNN as a core technique simultaneously aided by all of the three auxiliary techniques. They are presegmenting by the threshold T=173, preassigning a class number and providing teacher’s information. Therefore, it is feasible to apply the MBNN-3P to segment the G band images of T. monococcum chromosomes. This study might be of great significance to improve the accuracy and speed of automated analysis of plant chromosomes.

Introducing the theory of fuzzy set, mathematical morphology and computerized mask fast scanning, we developed the TOOTH.SCA software and method to analyze the effect of fluoride (NaF) on ore content of human tooth enamel automatically and quantitatively. And we obtained some characteristic parameters, such as the depth, the type and the demineralized content of every scathing layer of dental caries. The smallest scale of mask scanning is 0.1 microm x 0.1 microm and the time required to analyze a sample is only 12 s. The applied software and method we built play an important role to the research on the mechanism of pathological changes of teeth and preventing dental caries.

Wavelet neural network is a kind of neural networks, which can closely combine wavelet theory with neural network theory, and avoid the blindness of framework designs for BP neural networks and the problem of nonlinear optimizations, such as local optimization. So it can greatly simplify the training of neural networks. It has good abilities of function learning and dissemination with a vast range of prospects for application. The theories of wavelet transformation and multiresolution analysis are expounded, and then the mathematical model for the wavelet neural networks and its applications are introduced.

Some computer applications for cell characterization in medicine and biology, such as analysis of surface structure of cell wall-deficient EVC (El Tor Vibrio of Cholera), operate with cell samples taken from very small areas of interest. In order to perform texture characterization in such an application, only a few texture operators can be employed: the operators should be insensitive to noise and image distortion and be reliable in order to estimate texture quality from images. Therefore, we introduce wavelet theory and mathematical morphology to analyse the cellular surface micro-area image obtained by SEM (Scanning Electron Microscope). In order to describe the quality of surface structure of cell wall-deficient EVC, we propose a fully automatic computerized method. The image analysis process is carried out in two steps. In the first, we decompose the given image by dyadic wavelet transform and form an image approximation with higher resolution, by doing so, we perform edge detection of given images efficiently. In the second, we introduce many operations of mathematical morphology to obtain morphological quantitative parameters of surface structure of cell wall-deficient EVC. The obtained results prove that the method can eliminate noise, detect the edge and extract the feature parameters validly. In this work, we have built automatic analytic software named "EVC.CELL".

AIM Automatic measuring and quantitative analysis of form feature parameter of the liposome and emulsion image. METHODS On the basis of analyzing the form feature of liposome, some algorithms, such as the logical algorithms of mathematical morphologic, the grayscale-gradient co-occurrence matrix-based method and the image enhancement method of grayscale image, were used to construct the parameter of extracting method and software. RESULTS The noise is restrained effectively and the definition of image increases, implementing the automatic measuring of particle parameters. CONCLUSION The analytical result of bipolysaccharide liposome and some emulsions indicate that the method can definitely increase the speed and precision of the measurement.

应用自组织特征映射神经网络原理与方法,建立了双层Kohonen神经网络,其第一层是从二维图象平面到二维特征平面的映射,用于染色体高分辨率带纹的提取和带纹参数的计算,第二层是高维特征参数阵列到二维聚类平面的映射,用于同源染色体的自动配对和分类.应用结果表明,该方法能快速、准确地实现对栽培小麦染色体高分辨率G带带纹图象的特征参数自动提取和自动配对.

由中国生物物理学会主办的第三届全国现代生物物理技术学术讨论会,于2000年10月12日至15日在张家界市召开,出席会议的有来自全国各著名大专院校和科研机构的61名专家、教授和科学工作者.会议共收到论文62篇.大会特邀报告14篇,专题报告36篇.代表们针对当前国内与国际上生物、医学领域中,现代生物物理技术的某些热点和关键问题,以及最新的仪器工程技术和分析技术研究的成果和发展动态,进行了广泛深入的学术讨论与交流.代表们认识到:21世纪是生命科学的世纪,世界各国都在争夺生命科学技术的制高点,因此生命科学相关的仪器与分析技术亦成为十分重要的前沿研究领域.生命科学的基础研究只有在现代科学仪器和最新的理论和分析技术的支持下,才可能获得重大突破.代表们认为引进国际先进的技术方法和成果是必要的,但更重要的是敢于创新.本次学术讨论会的目的是促进现代科学仪器和生物物理技术、方法在生命科学领域中的应用,并推进生命科学仪器与分析技术的创新.