Zhao Dazhe

Automated Detection of Pulmonary Nodules in HRCT Images

Lung cancers death rates have been the main cause of cancer deaths in the world, early detection and treatment of lung cancer can greatly improve the survival rate of patient. This paper presents an automatic computer-aided detection (CAD) scheme that can identify the pulmonary nodule at an early stage from CT images. The work is separated to several steps: the segmentation of lung parenchyma, trachea and main airway bronchi elimination, the filter of nodule candidates, the detection of nodule candidates, the feature extraction and classification, three-dimensional visualization. The clinical testing for lots of lung cancer patients revealed that its high detection sensitivity of pulmonary nodules can meet basically the requirement of clinical diagnosis.

Computer-Aided Lung Nodule Detection Based on CT Images

Lung cancer is the most common fatal malignancy in both men and women, early detection and treatment of lung cancer can greatly improve the survival rate of patient. This paper presents an automatic computer-aided detection (CAD) scheme that can identify the lung nodule at an early stage from CT images. The work is separated to several steps: the segmentation of lung parenchyma, the detection of suspicious nodule candidates, the feature extraction and classification, use adaptive threshold segmentation, math morphologic, Gaussian filter, Hessian matrix algorithms. Though clinical trials, the computer-aided detection scheme can help physician improve the diagnosis efficiency.

GTP Security in 3G Core Network

Nowadays, 3G netowrk plays a very important role in mobile communication system. But the security concern of such network, especially the core network, is far from being satisfied. GPRS Tunnel Protocol(GTP), which is one of the key protocols in the core network, is quite vulnerable to attacks in the flat, full IP environment. Therefore solving such problem properly is very urgent and important for the operation of 3G network. In this paper, a security analysis of GTP is presented, a defense solution for these security threats and an event-based description language are also proposed. The experiment result shows the potential of our solution.

Full documents support with forms in workflow systems

Form is the information media in the application system and becomes very important in the daily business processing, especially in the workflow enviroment. In this paper, we analyze the form related concepts, elements and show how to implement a form system with a real form system, call UniEAP Form. To play a full power in application systems, we address how the form system can be leveraged in a workflow system in the aspects of integrated development environment, single place for user authorization, runtime interaction, and transaction consistency. Many applications currently running proved the validity and efficiency of our Form and the integrated approach with workflow.

Research and Implementation of A Computer-aided Diagnosis Algorithm Platform

The accuracy of medical imaging equipment examination depends on the quality of both the image acquisition and computer-aided diagnosis. Therefore the development of computer-aided diagnosis system has become the most important factor of the development of medical imaging equipment. The paper introduces a computer-aided diagnosis algorithm platform (CADAP), which includes medical image 3D visualization, image enhancement, image segmentation, region of interest detection, feature extract and classification algorithm, and it can increase researchers productivity and efficiency to develop more CAD system. Through integrating workflow into algorithm platform, the algorithm platform whose process can be defined and reengineered by its user easily is developed based on component and multi architecture.

An automatic partitioning method of CTA head-neck image

The segmentation of vessel in CTA is a challenging task mainly due to bone contact from different regions. The proposed method considers a spatial structure features based partition algorithm which can be separated into three sub-volumes. Firstly, a region growing method was brought forward to remove the background interference, and then added up the bone pixel number in the canter region of head image with specific rules to separate from the proximal layer and the middle layer. Finally, the structure feature of bone profile in neck image was used to separate from middle layer and the distal layer. The experiment results illustrate that the proposed method effectively separate the CTA head-neck image into three sub-volumes in terms of real-timeness, effectiveness and robustness.

Neural network classification algorithm based on feature space optimization in application of pulmonary nodules detection

This paper firstly preprocesses CT images using a dot enhance filter which can enhance some round-like nodule tissues and at the same time, restrain interference from tissues of other shapes(for example, linear vessels). Then, the paper adopts a feature space optimization design theory to select seven effective features from twelve original candidate features and regards the combinations of the seven features as input feature vectors of a classifier. Finally, the paper uses a BP neural network classifier to achieve pulmonary nodules classification. The experiment shows that the method presented here can effectively reduce false positive of nodule detection, obtaining a better classification result.

A Two-Stage Range-Free Localization Method for Wireless Sensor Networks

Range-free localization plays an important role in low-cost and large scale wireless sensor networks. Many existing range-free localization methods encounter high localization error, especially for the network with a coverage hole. One reason for high localization error is unreasonable distance estimation method. Another reason is that unknown nodes use the shortest distance which has large cumulative distance error to estimate their positions. In this paper, a two-stage centralized range-free localization algorithm (TCRL) is proposed. In the first stage, we design a novel rational distance estimation method to alleviate the distance estimation error between neighbor nodes based on the connectivity information and geometric features. In the second stage, a novel neighborhood function is derived from the estimated distances between neighbor nodes. Then a new localization strategy is proposed based on greedy idea. Finally, the proposed algorithm is compared with the same type algorithms in two network scenarios, namely, random deployment and random deployment with a coverage hole. The simulation results show that TCRL achieves more accurate and reliable results than most of existing range-free methods in the two network scenarios.

The Parametric Design Based on Bee Dual Population Evolution on Chaos

Geometric constraint problem is equivalent to the problem of solving a set of nonlinear equations substantially. We propose a new optimization algorithm—Bee Dual Population Evolution on Chaos Algorithm and apply it into the geometric constraint solving. By the use of properties of ergodicity, randomicity and regularity of chaos motion, genetic algorithm of bee dual population evolution based on chaos is proposed. On the basis of genetic algorithm based on bee dual population evolution, the algorithm adopts chaos optimization to improve species’ initial quality and utilizes the annealing chaotic mutation operator to replace the standard mutation operator in order to avoid the search being trapped in local optimum. The experiment shows that Bee Dual Population Evolution on Chaos has good capability in the geometric constraint solving.

A development of nonlinear transparency display algorithm

This paper introduces the simple liner transparent model and nonlinear transparency model which based on Z component. It analyses the disadvantage which existed in the Z component and brings forward an algorithm of nonlinear transparent display model which based on bounding box. The experiment shows that we can conquer the shortcoming which generated by nonlinear transparent model that based on Z component. This algorithm resolves the problem, which brings great variety on the grads and the edge of the contour becomes too bright, during the objects displaying. Then we can gain better 3D transparency display effect.