Shandong Wu

Chaotic invariants of Lagrangian particle trajectories for anomaly detection in crowded scenes

A novel method for crowd flow modeling and anomaly detection is proposed for both coherent and incoherent scenes. The novelty is revealed in three aspects. First, it is a unique utilization of particle trajectories for modeling crowded scenes, in which we propose new and efficient representative trajectories for modeling arbitrarily complicated crowd flows. Second, chaotic dynamics are introduced into the crowd context to characterize complicated crowd motions by regulating a set of chaotic invariant features, which are reliably computed and used for detecting anomalies. Third, a probabilistic framework for anomaly detection and localization is formulated. The overall work-flow begins with particle advection based on optical flow. Then particle trajectories are clustered to obtain representative trajectories for a crowd flow. Next, the chaotic dynamics of all representative trajectories are extracted and quantified using chaotic invariants known as maximal Lyapunov exponent and correlation dimension. Probabilistic model is learned from these chaotic feature set, and finally, a maximum likelihood estimation criterion is adopted to identify a query video of a scene as normal or abnormal. Furthermore, an effective anomaly localization algorithm is designed to locate the position and size of an anomaly. Experiments are conducted on known crowd data set, and results show that our method achieves higher accuracy in anomaly detection and can effectively localize anomalies.

Action recognition in videos acquired by a moving camera using motion decomposition of Lagrangian particle trajectories

Recognition of human actions in a video acquired by a moving camera typically requires standard preprocessing steps such as motion compensation, moving object detection and object tracking. The errors from the motion compensation step propagate to the object detection stage, resulting in miss-detections, which further complicates the tracking stage, resulting in cluttered and incorrect tracks. Therefore, action recognition from a moving camera is considered very challenging. In this paper, we propose a novel approach which does not follow the standard steps, and accordingly avoids the aforementioned difficulties. Our approach is based on Lagrangian particle trajectories which are a set of dense trajectories obtained by advecting optical flow over time, thus capturing the ensemble motions of a scene. This is done in frames of unaligned video, and no object detection is required. In order to handle the moving camera, we propose a novel approach based on low rank optimization, where we decompose the trajectories into their camera-induced and object-induced components. Having obtained the relevant object motion trajectories, we compute a compact set of chaotic invariant features which captures the characteristics of the trajectories. Consequently, a SVM is employed to learn and recognize the human actions using the computed motion features. We performed intensive experiments on multiple benchmark datasets and two new aerial datasets called ARG and APHill, and obtained promising results.

Flexible signature descriptions for adaptive motion trajectory representation, perception and recognition

Motion trajectory is a meaningful and informative clue in characterizing the motions of human, robots or moving objects. Hence, it is important to explore effective motion trajectory modeling. However, with the existing methods, a motion trajectory is used in its raw data form and effective trajectory description is lacking. In this paper, we propose a novel 3D motion trajectory signature descriptor and develop three signature descriptions for motion characterization. The flexible descriptions give the signature high functional adaptability to meet various application requirements in trajectory representation, perception and recognition. The full signature, optimized signature and cluster signature are firstly defined for trajectory representation. Then we explore the motion perception from a single signature, inter-signature matching and the generalization of a cluster signature. Furthermore, three solutions for signature recognition are investigated corresponding to different signature descriptions. The conducted experiments verified the signatures capabilities and flexibility. The signatures application to robot learning is also discussed.

Automated chest wall line detection for whole-breast segmentation in sagittal breast MR images.

PURPOSE Breast magnetic resonance imaging (MRI) plays an important role in the clinical management of breast cancer. Computerized analysis is increasingly used to quantify breast MRI features in applications such as computer-aided lesion detection and fibroglandular tissue estimation for breast cancer risk assessment. Automated segmentation of the whole-breast as an organ from the other parts imaged is an important step in aiding lesion localization and fibroglandular tissue quantification. For this task, identifying the chest wall line (CWL) is most challenging due to image contrast variations, intensity discontinuity, and bias field. METHODS In this work, the authors develop and validate a fully automated image processing algorithm for accurate delineation of the CWL in sagittal breast MRI. The CWL detection is based on an integrated scheme of edge extraction and CWL candidate evaluation. The edge extraction consists of applying edge-enhancing filters and an edge linking algorithm. Increased accuracy is achieved by the synergistic use of multiple image inputs for edge extraction, where multiple CWL candidates are evaluated by the dynamic time warping algorithm coupled with the construction of a CWL reference. Their method is quantitatively validated by a dataset of 60 3D bilateral sagittal breast MRI scans (in total 3360 2D MR slices) that span the full American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) breast density range. Agreement with manual segmentation obtained by an experienced breast imaging radiologist is assessed by both volumetric and boundary-based metrics, including four quantitative measures. RESULTS In terms of breast volume agreement with manual segmentation, the overlay percentage expressed by the Dices similarity coefficient is 95.0% and the difference percentage is 10.1%. More specifically, for the segmentation accuracy of the CWL boundary, the CWL overlay percentage is 92.7% and averaged deviation distance is 2.3 mm. Their method requires ≈ 4.5 min for segmenting each 3D breast MRI scan (56 slices) in comparison to ≈ 35 min required for manual segmentation. Further analysis indicates that the segmentation performance of their method is relatively stable across the different BI-RADS density categories and breast volume, and also robust with respect to a varying range of the major parameters of the algorithm. CONCLUSIONS Their fully automated method achieves high segmentation accuracy in a time-efficient manner. It could support large scale quantitative breast MRI analysis and holds the potential to become integrated into the clinical workflow for breast cancer clinical applications in the future.

On Signature Invariants for Effective Motion Trajectory Recognition

Motion trajectory can be an informative and descriptive clue that is suitable for the characterization of motion. Studying motion trajectory for effective motion description and recognition is important in many applications. For instance, motion trajectory can play an important role in the representation, recognition and learning of most long-term human or robot actions, behaviors and activities. However, effective trajectory descriptors are lacking and most reported work just uses motion trajectory in its raw data form. In this paper, we propose a novel motion trajectory signature descriptor and study its rich descriptive invariants which benefit effective motion trajectory recognition. These invariants are key measures of the flexibility and effectiveness of a descriptor. Substantial descriptive invariants can be deduced from the proposed trajectory signature, which is attributed to the computational locality of the signature components. We first present the signature definition and its robust implementation. Then the signatures invariants are elaborated. A non-linear inter-signature matching algorithm is developed to measure the signatures similarity for trajectory recognition. Experiments are conducted to recognize human sign language, in which both synthetic and real data are used to verify the signatures invariants, and to illustrate the effectiveness in the signature recognition.

Quantitative assessment of background parenchymal enhancement in breast MRI predicts response to risk-reducing salpingo-oophorectomy: preliminary evaluation in a cohort of BRCA1/2 mutation carriers

IntroductionWe present a fully automated method for deriving quantitative measures of background parenchymal enhancement (BPE) from breast dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) and perform a preliminary evaluation of these measures to assess the effect of risk-reducing salpingo-oophorectomy (RRSO) in a cohort of breast cancer susceptibility gene 1/2 (BRCA1/2) mutation carriers.MethodsBreast DCE-MRI data from 50 BRCA1/2 carriers were retrospectively analyzed in compliance with the Health Insurance Portability and Accountability Act and with institutional review board approval. Both the absolute (| |) and relative (%) measures of BPE and fibroglandular tissue (FGT) were computed from the MRI scans acquired before and after RRSO. These pre-RRSO and post-RRSO measures were compared using paired Student’s t test. The area under the curve (AUC) of the receiver operating characteristic (ROC) was used to evaluate the performance of relative changes in the BPE and FGT measures in predicting breast cancer that developed in these women after the RRSO surgery.ResultsFor the 44 women who did not develop breast cancer after RRSO, the absolute volume of BPE and FGT had a significant decrease (P < 0.05) post-RRSO, whereas for the 6 women who developed breast cancer, there were no significant changes in these measures. Higher values in all BPE and FGT measures were also observed post-RRSO for the women who developed breast cancer, compared with women who did not. Relative changes in BPE percentage were most predictive of women who developed breast cancer after RRSO (P < 0.05), whereas combining BPE percentage and |FGT| yielded an AUC of 0.80, higher than BPE percentage (AUC = 0.78) or |FGT| (AUC = 0.66) alone (both P > 0.02).ConclusionsQuantitative measures of BPE and FGT are different before and after RRSO, and their relative changes are associated with prediction of developing breast cancer, potentially indicative of women who are more susceptible to develop breast cancer after RRSO in BRCA1/2 mutation carriers.

Impact of emphysema heterogeneity on pulmonary function.

Objectives To investigate the association between emphysema heterogeneity in spatial distribution, pulmonary function and disease severity. Methods and Materials We ascertained a dataset of anonymized Computed Tomography (CT) examinations acquired on 565 participants in a COPD study. Subjects with chronic bronchitis (CB) and/or bronchodilator response were excluded resulting in 190 cases without COPD and 160 cases with COPD. Low attenuations areas (LAAs) (≤950 Hounsfield Unit (HU)) were identified and quantified at the level of individual lobes. Emphysema heterogeneity was defined in a manner that ranged in value from −100% to 100%. The association between emphysema heterogeneity and pulmonary function measures (e.g., FEV1% predicted, RV/TLC, and DLco% predicted) adjusted for age, sex, and smoking history (pack-years) was assessed using multiple linear regression analysis. Results The majority (128/160) of the subjects with COPD had a heterogeneity greater than zero. After adjusting for age, gender, smoking history, and extent of emphysema, heterogeneity in depicted disease in upper lobe dominant cases was positively associated with pulmonary function measures, such as FEV1 Predicted (p<.001) and FEV1/FVC (p<.001), as well as disease severity (p<0.05). We found a negative association between HI% , RV/TLC (p<0.001), and DLco% (albeit not a statistically significant one, p = 0.06) in this group of patients. Conclusion Subjects with more homogeneous distribution of emphysema and/or lower lung dominant emphysema tend to have worse pulmonary function.

Atlas-Based probabilistic fibroglandular tissue segmentation in breast MRI

In this paper we propose an atlas-aided probabilistic model-based segmentation method for estimating the fibroglandular tissue in breast MRI, where a novel fibroglandular tissue atlas is learned to aid the segmentation. The atlas represents a pixel-wise likelihood of being fibroglandular tissue in the breast, which is derived by combining deformable image warping, using aligned breast contour points as landmarks, with a kernel density estimation technique. A mixture multivariate model is learned to characterize the breast tissue using MR image features, and the segmentation is subsequently based on examining the posterior probability where the learned atlas is incorporated as the prior probability. In our experiments, the algorithm-generated segmentation results of 10 cases are compared to the manual segmentations, verified by an experienced breast imaging radiologist, to assess the accuracy of the algorithm, where the Dices Similarity Coefficient (DSC) shows a 0.85 agreement. The proposed automated segmentation method could be used to estimate the volumetric amount of fibroglandular tissue in the breast for breast cancer risk estimation.

Fully-automated fibroglandular tissue segmentation in breast MRI

We propose an automated segmentation method for estimating the fibroglandular (i.e., dense) tissue in breast MRI. The first step of our method is to segment the breast as an organ from other imaged parts through an integrated edge extraction and voting algorithm. Then, we apply the nonparametric non-uniform intensity normalization (N3) algorithm to the segmented breast to correct bias field which is common in breast MRI. After that, fuzzy C-means clustering is performed to categorize the breast tissue into two clusters, i.e., fibroglandular tissue and fat. The automated segmentation results are compared to manual segmentations, verified by an experienced breast imaging radiologist, to assess the accuracy of the algorithm, where the Dices Similarity Coefficient (DSC) shows a 0.73 agreement in our experiments. The benefit of the bias correction step is also shown through the comparison with the results obtained by excluding the bias correction step.

Visual tracking based on group sparsity learning

We propose a new tracking method based on a group sparsity learning model. Previous work on sparsity tracking rely on a single sparse model to characterize the templates of tracking targets, which is hard to express complex tracking scenes. In this work, we utilize a superposition of multiple simpler sparse models to capture the structural information across templates. More specifically, our tracking method is formulated within particle filter framework and the particle representations are decomposed into two sparsity norms: a