Shengdong Nie

University of Shanghai for Science and Technology, Shanghai, Shanghai Shi, China

Are you Shengdong Nie?

Claim your profile

Publications (6)0 Total impact

  • [Show abstract] [Hide abstract]
    ABSTRACT: A computerized scheme used for pulmonary parenchyma segmentation in CT images is developed based on improved fast marching and mathematical morphology methods. Firstly, apply the mathematical morphology to remove the noise and extract the gradient feature. And then, an improved fast marching method, in which speed function is confined by a threshold parameter, is used to segment pulmonary parenchyma in CT images. Experiment results showed that this method can segment pulmonary parenchyma correctly.
    Bioinformatics and Biomedical Engineering, 2008. ICBBE 2008. The 2nd International Conference on; 06/2008
  • Zehui Li, Shengdong Nie, Zhaoxue Chen
    [Show abstract] [Hide abstract]
    ABSTRACT: The performance of supervised learning classifier could be greatly increased by compressing redundant image features information. This paper proposed a new feature extraction algorithm using independent component analysis (ICA) for classification problems. Firstly extract original gray and texture image features (original features), then use ICA for obtaining independent components of the original features to compress redundant information, the new features were classified with support vector machines (SVM). The experiment results shows that the use of new features based on ICA greatly reduce the dimension of feature space and upgrade the performance of classifying systems. With the proposed ICA method, 2.17% higher accuracy was obtained than that of the original image features.
    01/2008;
  • Yingli Zhang, Shengdong Nie, Zhaoxue Chen, Wen Li
    [Show abstract] [Hide abstract]
    ABSTRACT: In the quantitative analysis of brain tissues (white matter, gray matter and cerebrospinal fluid) in magnetic resonance (MR) brain images, segmentation is the preliminary step. This paper proposes a novel segmentation method based on k-means objective function combined genetic algorithm, which is known for its global optimum searching ability. The method operates slice by slice via three main steps: (1) the non-brain tissues are removed from the original images using level set method, (2) the bias in the images which is caused by the inhomogeneity in the magnetic field is corrected by statistic method, and (3) the brain tissues are classified by k-means objective function combined genetic algorithm. The performance of the segmentation method was evaluated by the comparison with the fuzzy c-means (FCM) algorithm which is commonly used in segmentation of MR brain images. The accuracy of the proposed method is 3.21% higher than that of FCM algorithm. The time consumed in the proposed method is 0.596 second per image.
    Bioinformatics and Biomedical Engineering, 2007. ICBBE 2007. The 1st International Conference on; 08/2007
  • Wen Li, Yongfeng Huang, Xin Tian, Shengdong Nie
    [Show abstract] [Hide abstract]
    ABSTRACT: In this paper, we present an experimental research on the frameless registration of DSA/CT images based on frameless localization algorithm. The result shows that, 3D fusion and registration of vessels in the DSA images and anatomical structures in CT images will help surgeons to make accurate diagnosis and on plann operative.
    Sheng wu yi xue gong cheng xue za zhi = Journal of biomedical engineering = Shengwu yixue gongchengxue zazhi 03/2007; 24(1):23-5, 44.
  • Shengdong Nie, Yingli Zhang, Wen Li, Zhaoxue Chen
    [Show abstract] [Hide abstract]
    ABSTRACT: Image segmentation is the key step for quantitative analysis of brain tissues (white matter, gray matter and cerebrospinal fluid). Based on genetic algorithm and fuzzy C-means (FCM) approach, a fast and fully automatic segmentation method of brain tissues named genetic fuzzy clustering algorithm is introduced in this paper. The method operates slice by slice based on three main steps: The non-brain tissues are removed from the original head MR images at first using an auto-threshold method; then the initial cluster centers of FCM are determined by genetic algorithm; and finally brain tissues are segmented into white matter, grey matter and cerebrospinal fluid by FCM via only one iteration computation. The experiment results have shown that the segmentation method proposed by this paper has faster speed and higher accuracy compared with fast fuzzy c-means algorithm which is commonly used in segmentation of brain tissues.
    Conference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 02/2007; 2007:5628-33.
  • Zhaoxue Chen, Xiwen Sun, Shengdong Nie
    [Show abstract] [Hide abstract]
    ABSTRACT: Based on special distributing characteristics of pixel intensity in lung CT images, an efficient lung segmentation method is introduced. Associating approach of image threshold with fast region flood filling technique, this method can extract pulmonary parenchyma from CT images simply. After a preprocessing step for noise removal, it segments the lung CT image slice utilizing a threshold method at first, and then applies a fast and simple method to finish flood filling of the non-lung area. In the following steps, the lung area can be extracted automatically after an erosion operation and an area-filtering step. The presented experiment results have proved its validity.
    Conference proceedings: ... Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society. Conference 02/2007; 2007:5540-2.