Key Laboratory of Medical Image Computing of Ministry of Education

Shenyang, China

Key Laboratory of Medical Image Computing of Ministry of Education

Shenyang, China
SEARCH FILTERS
Time filter
Source Type

Jiang H.,Northeastern University China | Jiang H.,Key Laboratory of Medical Image Computing of Ministry of Education | Zhang Y.,Northeastern University China | Jiang J.,Northeastern University China | And 3 more authors.
Journal of Information and Computational Science | Year: 2014

Aiming at the long running time problem of the super-resolution image reconstruction method based on learning, this paper proposes a novel fast method. Principal Component Analysis (PCA) is used to reduce the data dimensionality of training data set and Vector Quantization (VQ) is introduced into super-resolution image reconstruction to divide subset. Both accelerate running of the method speed and solve the long running time problem caused by large amount of training data. In order to ensure the quality of the output image, Stationary Wavelet Transform (SWT) is used to extract the low and high frequency information of the sample image. And Markov Network is improved to find the best candidate block. Experiments show that without sacrificing the quality of final output high resolution image, the execution speed of the proposed method is greatly improved. 1548-7741/Copyright © 2014 Binary Information Press.


Jiang H.,Northeastern University China | Jiang H.,Key Laboratory of Medical Image Computing of Ministry of Education | Zheng R.,Northeastern University China | Yi D.,Shenyang University | Zhao D.,Northeastern University China
Computational and Mathematical Methods in Medicine | Year: 2013

A novel multi-instance learning (MIL) method is proposed to recognize liver cancer with abdominal CT images based on instance optimization (IO) and support vector machine with parameters optimized by a combination algorithm of particle swarm optimization and local optimization (CPSO-SVM). Introducing MIL into liver cancer recognition can solve the problem of multiple regions of interest classification. The images we use in the experiments are liver CT images extracted from abdominal CT images. The proposed method consists of two main steps: (1) obtaining the key instances through IO by texture features and a classification threshold in classification of instances with CPSO-SVM and (2) predicting unknown samples with the key instances and the classification threshold. By extracting the instances equally based on the entire image, the proposed method can ignore the procedure of tumor region segmentation and lower the demand of segmentation accuracy of liver region. The normal SVM method and two MIL algorithms, Citation-kNN algorithm and WEMISVM algorithm, have been chosen as comparing algorithms. The experimental results show that the proposed method can effectively recognize liver cancer images from two kinds of cancer CT images and greatly improve the recognition accuracy. © 2013 Huiyan Jiang et al.


Jiang H.,Northeastern University China | Jiang H.,Key Laboratory of Medical Image Computing of Ministry of Education | He B.,Northeastern University China | Fang D.,Northeastern University China | And 3 more authors.
Computational and Mathematical Methods in Medicine | Year: 2013

We propose a region growing vessel segmentation algorithm based on spectrum information. First, the algorithm does Fourier transform on the region of interest containing vascular structures to obtain its spectrum information, according to which its primary feature direction will be extracted. Then combined edge information with primary feature direction computes the vascular structure's center points as the seed points of region growing segmentation. At last, the improved region growing method with branch-based growth strategy is used to segment the vessels. To prove the effectiveness of our algorithm, we use the retinal and abdomen liver vascular CT images to do experiments. The results show that the proposed vessel segmentation algorithm can not only extract the high quality target vessel region, but also can effectively reduce the manual intervention. © 2013 Huiyan Jiang et al.


Jiang H.,Northeastern University China | Jiang H.,Key Laboratory of Medical Image Computing of Ministry of Education | Shi S.,Northeastern University China | Zhang Y.,Northeastern University China | And 5 more authors.
Journal of Information and Computational Science | Year: 2014

This paper proposed a novel image compression algorithm based on Gabor transform and fractal theory. First, texture features of image blocks were extracted based on Gabor transform; then cluster process were conducted for the R block and D block based on K-means algorithm, and in the process of block match, only those blocks belong to the same kind were taken into account, which can effectively reduce the encoding time. During the construction of the candidate blocks pool, two simplified approaches were introduced, including the simplification of the eight equidistant transforms and the outliers removed algorithm. Experimental results show that the proposed algorithm can both effectively speed up the encoding and ensure the image perfect visual effect, and has a good robustness. © 2014 by Binary Information Press.


Jiang H.,Northeastern University China | Jiang H.,Key Laboratory of Medical Image Computing of Ministry of Education | Lou B.,Northeastern University China | Zhang B.,Northeastern University China
Journal of Information and Computational Science | Year: 2012

Aiming at existing color image enhancement method cannot maintain the overall color of the image is bright and colorful and over-enhanced cause distortion when contrast is raised, this paper proposed enhancement algorithms based on Retinex-LIP combining the LIP Theory and the Retinex theory. First, through the Retinex adaptive luminance component of the color image process to solve color distortion and halos; and then under LIP theoretical framework, color image enhancement algorithm based on the recursive algorithm combined with Optimize the brightness stretching algorithm to compensate for color. Experimental results show that this method can provide the whit region of the original image from being stretched too bright. And it can also compensate the overall color of the image and highlight the details of the image's bright area and shaded area. In this way, the algorithm makes the image more vivid and clear. Copyright © 2012 Binary Information Press.


Jiang H.,Northeastern University China | Jiang H.,Key Laboratory of Medical Image Computing of Ministry of Education | Liu J.,Northeastern University China
Journal of Information and Computational Science | Year: 2012

In this paper, a new method is proposed to automatically diagnose early cirrhosis of liver in CT images. It is developed for classifying the normal liver and cirrhosis based on an auto-recognition method through two steps. Firstly, the Statistical Shape Model (SSM) is constructed to quantitatively evaluate the variations of the liver shape. The shape model of segmented liver can be presented by the mean shape and a number of modes of variation. Secondly, the classifier Support Vector Machine (SVM) optimized by Ant Colony Optimization (ACO) is used to classify the normal liver and cirrhosis. The quantitative evaluation of the proposed method shows that it can recognize liver with high accuracy whether it is normal or abnormal. The process of the method could be developed for diagnosis of any shape-related liver disease. Copyright © 2012 Binary Information Press.


Jiang H.,Northeastern University China | Jiang H.,Key Laboratory of Medical Image Computing of Ministry of Education | Fang D.,Northeastern University China | Gao R.,Northeastern University China
Journal of Information and Computational Science | Year: 2012

In this paper, we propose a novel coded aperture and two associated algorithm for making up the defect of traditional aperture such as the depth of scene, quality and the limitation of dimension. In sampling, combining with the sampling characteristics of traditional aperture and the properties of wavelet transform, we design a coded aperture achieving multi-angle reuse acquisition of light field. Then we project the captured light field data to double tree discrete wavelet domain. For reconstruction, we propose to use the sparse characteristics of light field combine with the redundancy in both spatial and angle domain. Experimental results demonstrate the efficiency and quality of the proposed methods. © 2012 Binary Information Press.

Loading Key Laboratory of Medical Image Computing of Ministry of Education collaborators
Loading Key Laboratory of Medical Image Computing of Ministry of Education collaborators