Key Laboratory of Medical Image Computing

Laboratory of, China

Key Laboratory of Medical Image Computing

Laboratory of, China
SEARCH FILTERS
Time filter
Source Type

Qu W.,Northeastern University China | Wang D.,Northeastern University China | Wang D.,Key Laboratory of Medical Image Computing | Feng S.,Northeastern University China | And 5 more authors.
Science China Information Sciences | Year: 2017

With the growing popularity of multimodal data on the Web, cross-modal retrieval on large-scale multimedia databases has become an important research topic. Cross-modal retrieval methods based on hashing assume that there is a latent space shared by multimodal features. To model the relationship among heterogeneous data, most existing methods embed the data into a joint abstraction space by linear projections. However, these approaches are sensitive to noise in the data and are unable to make use of unlabeled data and multi-modal data with missing values in real-world applications. To address these challenges, we proposed a novel multimodal deep-learning-based hash (MDLH) algorithm. In particular, MDLH uses a deep neural network to encode heterogeneous features into a compact common representation and learns the hash functions based on the common representation. The parameters of the whole model are fine-tuned in a supervised training stage. Experiments on two standard datasets show that the method achieves more effective results than other methods in cross-modal retrieval. © 2017, Science China Press and Springer-Verlag Berlin Heidelberg.


Kang H.,Northeastern University China | Liu J.,Northeastern University China | Xu L.,Northeastern University China | Xu L.,Key Laboratory of Medical Image Computing | And 4 more authors.
2016 IEEE International Conference on Information and Automation, IEEE ICIA 2016 | Year: 2016

The paper proposes a mathematical model about the relationship between the curvature radius changes of interventricular septal configuration and the pressure difference of left and right ventricular based on Young-Laplace model. The method of noninvasive detection of pressure difference of left and right ventricular is discussed. In this paper a ventricular model which can simulate the deformation of interventricular septal configuration in differential pressure is designed and produced. After scanning the ventricular model by CT, the boundary of the ventricular model is extracted and the curvature radius of the boundary is calculated. Software ANSYS is used to simulate the curvature of the ventricular septum model and verify the experiment. Then a relationship between the deformation of ventricular septum and the pressure difference of ventricle was established by the Hooke's law combined with the modified Young-Laplace model. After the verification, the modified Young-Laplace model can be applied to estimate pressure difference in ventricle. This paper contributes the development of noninvasive medical diagnosis of abnormal ventricular pressure. © 2016 IEEE.


Jiang H.,Northeastern University China | Jiang H.,Key Laboratory of Medical Image Computing | Zhao D.,Northeastern University China | Feng T.,Northeastern University China | And 2 more authors.
Computational and Mathematical Methods in Medicine | Year: 2013

A novel method is proposed to establish the classifier which can classify the pancreatic images into normal or abnormal. Firstly, the brightness feature is used to construct high-order tensors, then using multilinear principal component analysis (MPCA) extracts the eigentensors, and finally, the classifier is constructed based on support vector machine (SVM) and the classifier parameters are optimized with quantum simulated annealing algorithm (QSA). In order to verify the effectiveness of the proposed algorithm, the normal SVM method has been chosen as comparing algorithm. The experimental results show that the proposed method can effectively extract the eigenfeatures and improve the classification accuracy of pancreatic images. © 2013 Huiyan Jiang et al.


Jiang H.,Northeastern University China | Jiang H.,Key Laboratory of Medical Image Computing | Ma Z.,Northeastern University China | Hu Y.,Northeastern University China | And 2 more authors.
Computational Intelligence and Neuroscience | Year: 2012

An optimized medical image compression algorithm based on wavelet transform and improved vector quantization is introduced. The goal of the proposed method is to maintain the diagnostic-related information of the medical image at a high compression ratio. Wavelet transformation was first applied to the image. For the lowest-frequency subband of wavelet coefficients, a lossless compression method was exploited; for each of the high-frequency subbands, an optimized vector quantization with variable block size was implemented. In the novel vector quantization method, local fractal dimension (LFD) was used to analyze the local complexity of each wavelet coefficients, subband. Then an optimal quadtree method was employed to partition each wavelet coefficients, subband into several sizes of subblocks. After that, a modified K-means approach which is based on energy function was used in the codebook training phase. At last, vector quantization coding was implemented in different types of sub-blocks. In order to verify the effectiveness of the proposed algorithm, JPEG, JPEG2000, and fractal coding approach were chosen as contrast algorithms. Experimental results show that the proposed method can improve the compression performance and can achieve a balance between the compression ratio and the image visual quality. © 2012 Huiyan Jiang et al.


Xu L.,Northeastern University China | Xu L.,Key Laboratory of Medical Image Computing | He D.,Northeastern University China | Zhao Y.,Northeastern University China | Yin S.,Northeastern University China
Recent Patents on Biomedical Engineering | Year: 2013

Compared with the traditional method of measuring the blood pressure, the cuff-less blood pressure measuring method can detect the blood pressure continuously in a comfortable way, in order to provide an adequate basis of clinical diagnosis and has great significance to clinical and medical research. Some reported non-invasive continuous blood pressure measurements are presented in this review paper comprising illustrations and comparisons of the characteristic of the patents based on different methods, such as arterial tonometry method, vascular unloading technique method, pulse wave transit time method and others. This review also presents the characteristics and application of some related products. The challenge and the prospective of non-invasive continuous blood pressure measurement techniques are also analyzed. Finally, to conclude, non-invasive continuous blood pressure measurement is becoming the future development trend of blood pressure measurement, however, the accuracy of measurement also needs to be improved. © 2014 Bentham Science Publishers.


Yao Y.,Northeastern University China | Hao L.,Liaoning Medical University | Geng N.,Northeastern University China | Jin Y.,Northeastern University China | And 3 more authors.
Proceedings of the World Congress on Intelligent Control and Automation (WCICA) | Year: 2015

Carotid pressure waveform is often used to substitute the central aortic pressure waveform, which conveys lots of information regarding cardiovascular system. Thus, a method was proposed to reconstruct the carotid pressure waveform from the radial pressure waveform which can be more conveniently monitored in comparison with the carotid pressure waveform. The reconstruction method was using Finite Impulse Response (FIR) model to calculate mathematical transfer function (TF) between the radial pressure waveform and the carotid pressure waveform. Pulse pressure waveforms of 5 subjects were recorded to test the performance of the TF. Except for some details, the reconstructed carotid pressure waveform is acceptable with the best percent root-mean-square difference (PRD) of 13.60%. © 2014 IEEE.


Zhang S.,Northeastern University China | Chen Z.,Northeastern University China | Gu S.,Northeastern University China | Qiu X.,Northeastern University China | And 3 more authors.
ICMIPE 2013 - Proceedings of 2013 IEEE International Conference on Medical Imaging Physics and Engineering | Year: 2013

Breast tumor detection is a most effective way to immunized against mammary cancer. It is known that the sort algorithm of extreme learning machine(ELM), in view of the feature model for breast X-ray image, is being applied in the computer aided detection of breast masses. On the basis of all these, it is raised in this paper that marking for the suspicious region in the double view mammography by the use of ELM, then classifying the result of double views marking by using the Simple Bias classifier and finally gaining the detection result. The experiment with 444 cases or 222 pair of X-ray mammography from Liao Ning Province Cancer Hospital shows that, the breast tumor detection in double views mammography based on Simple Bias is an available and effective way to detect breast tumor. Key Words: Extreme learning machine, Simple Bias, mammography, double views, tumor detection. © 2013 IEEE.


Wang Z.,Northeastern University China | Wang Z.,Key Laboratory of Medical Image Computing | Qu Q.,Northeastern University China | Yu G.,Northeastern University China | Kang Y.,Northeastern University China
Neural Computing and Applications | Year: 2014

Mammography is one of the most important methods for breast tumor detection, while existing computer-aided diagnosis (CAD) technology based on single-view mammograms ignores the contrastive feature between medio-lateral oblique (MLO) and cranio-caudal (CC) views, and CAD technology based on double-view overlooks features of single views. But in clinical environment, radiologists not only read both CC view images and MLO view images individually, but also contrast these two types of views to diagnose each case. Therefore, to simulate diagnosis process of radiologists, in this paper, a fused feature model which blends features of single views with contrastive features of double views is proposed. The fused feature model is optimized by means of feature selection methods. Then, a CAD detection method based on extreme learning machine, a classifier with wonderful universal approximation capability, is proposed to improve the effectiveness of breast tumor detection by applying the optimum fused feature. The effectiveness of proposed method is verified by 222 pairs of mammograms from 222 women in Northeast China through the complete experiment. © 2014 The Natural Computing Applications Forum


Jia T.,Northeastern University China | Jia T.,Key Laboratory of Medical Image Computing | Zhang H.,Northeastern University China | Meng H.,Northeastern University China
Bio-Medical Materials and Engineering | Year: 2014

Lung vessels often interfere with the detection of lung nodules. In this paper, a novel computer-Aided lung nodule detection scheme on vessel segmentation is proposed. This paper describes an active contour model which can combine image region mean gray value and image edge energy. It is used to segment and remove lung vessels. A selective shape filter based on Hessian Matrix is used to detect suspicious nodules and remove omitted lung vessels. This paper extracts density, shape and position features of suspicious nodules, and uses a Rule-Based Classification (RBC) method to identify true positive nodules. In the experiment results, the detection sensitivity is about 90% and FP is 1/scan. © 2014 - IOS Press and the authors.


Yu E.,Northeastern University China | He D.,Northeastern University China | Su Y.,Northeastern University China | Zheng L.,Northeastern University China | And 3 more authors.
2013 IEEE International Conference on Robotics and Biomimetics, ROBIO 2013 | Year: 2013

Objective: To judge whether the pulse rate variability can be used as a surrogate of heart rate variability, as well as investigate the quantitative relationship between them. Methods: Being simultaneously acquired, the pulse wave and ECG data were denoised, removed baseline drift. Then the pulse rate intervals and heart rate intervals were extracted. Finally, the relationship between the heart rate variability and pulse rate variability were studied in the time domain, frequency domain, and nonlinear analysis. Conclusion: By studying the pulse rate variabilities and heart rate variabilities of 30 healthy adolescents, we find that heart rate variability and pulse rate variability is correlated, but the difference is relatively small, in a resting condition, the difference of time domain is less than 3%, the Frequency domain is less than 9%, the nonlinear is less than 9%, which in a certain extent can replace each other. Furthermore, the influence from the neural regulation and respiration caused the delay of PRV in comparison with HRV, ranging from 6% to 20% of a heartbeat period. The breathing more greatly influences the PRV than HRV. © 2013 IEEE.

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