Enhanced parameter estimation methods for noisy SPECT data

Functional imaging with PET and SPECT is capable of visualizing subtle changes in physiological function in vivo, which aids in the early diagnosis of disease. Quantitative functional parameters are usually derived by curve fitting the dynamic data of a functional imaging study. However, the intrinsic high level of noise and low signal to noise ratio can lead to instability in the parameter estimation and give rise to non-physiological parameter estimates. Clustering techniques have been applied to improve signal to noise ratio and the reliability of parametric image generation, but these may enhance partial volume effects (PVE) and result in biased estimates for small structures. Therefore, a systematic study was performed using computer simulations of SPECT data and the generalized linear least square algorithm (GLLS) to evaluate the ability of three proposed enhanced methods and a clustering-aided method to improve the reliability of parametric image generation. The results demonstrate that clustering with sufficient cluster numbers ameliorated PVE and provided noise-insensitive parameter estimates. The enhanced GLLS method with a prior volume of distribution and bootstrap Monte Carlo resampling improved the reliability of the curve fitting, and is thus suitable for application to noisy SPECT data.     

Automated PET-guided liver segmentation from low-contrast CT volumes using probabilistic atlas

The use of the functional PET information from PET-CT scans to improve liver segmentation from low-contrast CT data is yet to be fully explored. In this paper, we fully utilize PET information to tackle challenging liver segmentation issues including (1) the separation and removal of the surrounding muscles from liver region of interest (ROI), (2) better localization and mapping of the probabilistic atlas onto the low-contrast CT for a more accurate tissue classification, and (3) an improved initial estimation of the liver ROI to speed up the convergence of the expectation-maximization (EM) algorithm for the Gaussian distribution mixture model under the guidance of a probabilistic atlas. The primary liver extraction from the PET volume provides a simple mechanism to avoid the complicated pre-processing of feature extraction as used in the existing liver CT segmentation methods. It is able to guide the probabilistic atlas to better conform to the CT liver region and hence helps to overcome the challenge posed by liver shape variability. Our proposed method was evaluated against manual segmentation by experienced radiologists. Experimental results on 35 clinical PET-CT studies demonstrated that our method is accurate and robust in automated normal liver segmentation.     

Detecting unattended packages through human activity recognition and object association

This paper provides a novel approach to detect unattended packages in public venues. Different from previous works on this topic which are mostly limited to detecting static objects where no human is nearby, we provide a solution which can detect an unattended package with people in its close proximity but not its owners. Mimicking the human logic in detecting such an event, our decision-making is based on understanding human activity and the relationships between humans and packages. There are three main contributions from this paper. First, an efficient method is provided to online categorize moving objects into the predefined classes using the eigen-features and the support vector machines (SVM). Second, utilizing the classification results, a method is developed to recognize human activities with hidden Markov models (HMM) and decide the package ownership. Finally the unattended package detection is achieved by analyzing multiple object relationships: package ownership, spatial and temporal distance relationships.     

Performance evaluation of functional medical imaging compression via optimal sampling schedule designs and cluster analysis

In previous work we have described a technique for the compression of positron emission tomography (PET) image data in the spatial and temporal domains based on optimal sampling schedule designs (OSS) and cluster analysis. It can potentially achieve a high data compression ratio greater than 80:1. However, the number of distinguishable cluster groups in dynamic PET image data is a critical issue for this algorithm that has not been experimentally analyzed on clinical data. In this paper, the problem of experimentally determining the ideal cluster number for the algorithm for PET brain data is addressed.     

Segmentation of VOI from multidimensional dynamic PET images by integrating spatial and temporal features.

Segmentation of multidimensional dynamic positron emission tomography (PET) images into volumes of interest (VOIs) exhibiting similar temporal behavior and spatial features is a challenging task due to inherently poor signal-to-noise ratio and spatial resolution. In this study, we propose VOI segmentation of dynamic PET images by utilizing both the three-dimensional (3-D) spatial and temporal domain information in a hybrid technique that integrates two independent segmentation techniques of cluster analysis and region growing. The proposed technique starts with a cluster analysis that partitions the image based on temporal similarities. The resulting temporal partitions, together with the 3-D spatial information are utilized in the region growing segmentation. The technique was evaluated with dynamic 2-[18F] fluoro-2-deoxy-D-glucose PET simulations and clinical studies of the human brain and compared with the k-means and fuzzy c-means cluster analysis segmentation methods. The quantitative evaluation with simulated images demonstrated that the proposed technique can segment the dynamic PET images into VOIs of different kinetic structures and outperforms the cluster analysis approaches with notable improvements in the smoothness of the segmented VOIs with fewer disconnected or spurious segmentation clusters. In clinical studies, the hybrid technique was only superior to the other techniques in segmenting the white matter. In the gray matter segmentation, the other technique tended to perform slightly better than the hybrid technique, but the differences did not reach significance. The hybrid technique generally formed smoother VOIs with better separation of the background. Overall, the proposed technique demonstrated potential usefulness in the diagnosis and evaluation of dynamic PET neurological imaging studies.     

Geometric Analysis of Particle Motion in a Vector Image Field

This paper proposes a geometrical model for the Particle Motion in a Vector Image Field (PMVIF) method. The model introduces a c-evolute to approximate the edge curve in the gray-level image. The c-evolute concept has three major novelties: (1) The locus of Particle Motion in a Vector Image Field (PMVIF) is a c-evolute of image edge curve; (2) A geometrical interpretation is given to the setting of the parameters for the method based on the PMVIF; (3) The gap between the image edge’s critical property and the particle motion equations appeared in PMVIF is padded. Our experimental simulation based on the image gradient field is simple in computing and robust, and can perform well even in situations where high curvature exists. Chenggang Lu received his Bachelor of Science and PhD degrees from Zhejiang University in 1996 and 2003, respectively. Since 2003, he has been with VIA Software (Hang Zhou), Inc. and Huawei Technology, Inc. His research interests include image processing, acoustic signaling processing, and communication engineering. Zheru Chi received his BEng and MEng degrees from Zhejiang University in 1982 and 1985 respectively, and his PhD degree from the University of Sydney in March 1994, all in electrical engineering. Between 1985 and 1989, he was on the Faculty of the Department of Scientific Instruments at Zhejiang University. He worked as a Senior Research Assistant/Research Fellow in the Laboratory for Imaging Science and Engineering at the University of Sydney from April 1993 to January 1995. Since February 1995, he has been with the Hong Kong Polytechnic University, where he is now an Associate Professor in the Department of Electronic and Information Engineering. Since 1997, he has served on the organization or program committees for a number of international conferences. His research interests include image processing, pattern recognition, and computational intelligence. Dr. Chi has authored/co-authored one book and nine book chapters, and published more than 140 technical papers. Gang Chen received his Bachelor of Science degree from Anqing Teachers College in 1983 and his PhD degree in the Department of Applied Mathematics at Zhejiang University in 1994. Between 1994 and 1996, he was a postdoctoral researcher in electrical engineering at Zhejiang University. From 1997 to 1999, he was a visiting researcher in the Institute of Mathematics at the Chinese University of Hong Kong and the Department of Electronic and Information Engineering at The Hong Kong Polytechnic University. Since 2001, he has been a Professor at Zhejiang University. He has been the Director of the Institute of DSP and Software Techniques at Ningbo University since 2002. His research interests include applied mathematics, image processing, fractal geometry, wavelet analysis and computer graphics. Prof. Chen has co-authored one book, co-edited five technical proceedings and published more than 80 technical papers. (David) Dagan Feng received his ME in Electrical Engineering & Computing Science (EECS) from Shanghai JiaoTong University in 1982, MSc in Biocybernetics and Ph.D in Computer Science from the University of California, Los Angeles (UCLA) in 1985 and 1988 respectively. After briefly working as Assistant Professor at the University of California, Riverside, he joined the University of Sydney at the end of 1988, as Lecturer, Senior Lecturer, Reader, Professor and Head of Department of Computer Science/School of Information Technologies, and Associate Dean of Faculty of Science. He is Chair-Professor of Information Technology, Hong Kong Polytechnic University; Honorary Research Consultant, Royal Prince Alfred Hospital, the largest hospital in Australia; Advisory Professor, Shanghai JiaoTong University; Guest Professor, Northwestern Polytechnic University, Northeastern University and Tsinghua University. His research area is Biomedical & Multimedia Information Technology (BMIT). He is the Founder and Director of the BMIT Research Group. He has published over 400 scholarly research papers, pioneered several new research directions, made a number of landmark contributions in his field with significant scientific impact and social benefit, and received the Crump Prize for Excellence in Medical Engineering from USA. More importantly, however, is that many of his research results have been translated into solutions to real-life problems and have made tremendous improvements to the quality of life worldwide. He is a Fellow of ACS, HKIE, IEE, IEEE, and ATSE, Special Area Editor of IEEE Transactions on Information Technology in Biomedicine, and is the current Chairman of IFAC-TC-BIOMED.     

Texture image classification with discriminative neural networks

Computational Visual Media - Texture provides an important cue for many computer vision applications, and texture image classification has been an active research area over the past years....