My publications

Andriy Fedorov, Xiaoxing Li, Kilian M. Pohl, Sylvain Bouix, Martin Styner, Merideth Addicott, Chris Wyatt, James B. Daunais, William M. Wells and Ron Kikinis, Atlas-guided Segmentation of VervetMonkey BrainMRI, accepted by Special Issue on Neuroimaging of non-human Primates in Open Neuroimaging Journal (ONIJ), Jan. 2011. Abstract:The vervet monkey is an important nonhuman primate model that allows the study of isolated environmental factors in a controlled environment. Monkey MRI often suffers from lower quality images compared with human MRI because clinical equipment is typically used to image the smaller monkey brain and higher spatial resolution is required. This, together with the anatomical differences of the monkey brains, complicates the use of neuroimage analysis pipelines tuned for human MRI analysis. In this paper we developed an open source image analysis framework based on the tools available within the 3D Slicer software to support a biological study that investigates the effect of chronic ethanol exposure on brain morphometry in a longitudinally followed population of male vervets. We first developed a computerized atlas of vervet monkey brain MRI, which was used to encode the typical appearance of the individual brain structures in MRI and their spatial distribution. The atlas was then used as a spatial prior during automatic segmentation to process two longitudinal scans per subject. Our evaluation confirms the consistency and reliability of the automatic segmentation. The comparison of atlas construction strategies reveals that the use of a population-specific atlas leads to improved accuracy of the segmentation for subcortical brain structures. At the same time, the choice of atlas did not have a significant influence on the measured volumetric changes in the vervet brains between the acquisitions. The contribution of this work is twofold. First, we describe an image processing workflow specifically tuned towards the analysis of vervet MRI that consists solely of the open source software tools. Second, we develop a digital atlas of vervet monkey brain MRIs to enable similar studies that rely on the vervet model.

Xiaoxing Li, Xiaojing Long and Chris Wyatt, Registration of Images with Topological Change via Riemannian Embedding. IEEE International Symposium on Biomedical Imaging (ISBI), 2011. [PDF] Abstract:In this paper, we develop a new deformable registration algorithm for images with pathology-induced topological changes. In this algorithm, 3D images are embedded as 4D surfaces in a Riemannian space and the registration is conducted as a surface evolution process. Our algorithm differs from existing methods in the sense that it takes an a-priori estimation of areas with topological change as an additional input and generates dense deformation vector fields which are free of false deformation. In particular, the output of our algorithm is composed of a diffeomorphic deformation field and an intensity displacement which corrects the intensity difference caused by the topological changes. The experiments demonstrate that our proposed algorithm is capable of accurately registering images with considerable topological changes. More importantly, the resulting deformation field is not impacted by topological changes, i.e., there is no false deformation.

Xiaoxing Li and Chris Wyatt, Modeling topological changes in deformable registration, IEEE International Symposium on Biomedical Imaging (ISBI), Rotterdam, the Netherlands, April, 2010. [PDF] Abstract:Topological changes are usually seen in brain MR Images in aging or disease studies. For deformable registration algorithms, topological changes can cause false deformation in the resulting vector field, and affect algorithm convergence. In this work, we focus on the effect of topological changes on deformable registration algorithms that are inverse-consistent and diffeomorphic, e.g., diffeomorphic demons and symmetric LDDMM. We first use a simple example to demonstrate the adverse effect of topological changes on these algorithms. Then, we propose a novel framework that can be imposed onto any PDE-driven deformable registration algorithms, which renders them robust to topological changes. The output of our proposed framework consists of two components. The first is a deformation field that presents only the brain structural change which is the expected vector field if the topological change did not exist. The second component is a label map that provides a segmentation of the topological changes appeared in input images.

Xiaoxing Li and Chris Wyatt, Brain Segmentation Performance using T1-weighted Images versus T1 Maps. Proceedings of SPIE medical imaging, 2010. [PDF] Abstract:The recent driven equilibrium single-pulse observation of T1 (DESPOT1) approach permits real-time clinical acquisition of large-volume and high-isotropic-resolution T1 mapping of MR tissue parameters with improved uniformity. It is assumed that the quantitative nature of maps will facilitate clinical applications such as disvase diagnosis and comparison across subjects. However, there is not yet enough quantitative evidence on the actual benefit of adopting T1 maps, especially in computer-aided medical image analysis tasks. In this study, we compare methods with respect to image types, T1-weighted images or T1 maps, in automatic brain MRI segmentation. Our experimental results demonstrate that, using T1 maps, different segmentation algorithms show better agreement with each other, compared to that from using T1-weighted images. Furthermore, through multi-dimensional-scaling projection, we are able to visualize the relative affinity among segmentation results, which reveals that the projections of those segmentations using two different types of input images tend to form two separate clusters. Finally, by comparing to expert segmented reference segmentation of brain sub-regions, our results clearly indicate a better agreement between the manual reference and those automatic ones on T1 maps. In other words, our study provides an evidence for the hypothesis that compared to the conventionally used T1-weighted images, T1 maps lead to improved reliability in automatic brain MRI segmentation task.

Xiaoxing Li, Tao Jia and Hao Zhang, Expression-Insensitive 3D Face Recognition using Sparse Representation, Proceedings of IEEE Computer Vision and Pattern Recognition (CVPR), pp. 2575-2582, 2009. [PDF] Abstract:We present a face recognition method based on sparse representation for recognizing 3D face meshes under expressions using low-level geometric features. First, to enable the application of the sparse representation framework, we develop a uniform remeshing scheme to establish a consistent sampling pattern across 3D faces. To handle facial expressions, we design a feature pooling and ranking scheme to collect various types of low-level geometric features and rank them according to their sensitivities to facial expressions. By simply applying the sparse representation framework to the collected low-level features, our proposed method already achieves satisfactory recognition rates, which demonstrates the efficacy of the framework for 3D face recognition. To further improve results in the presence of severe facial expressions, we show that by choosing higher-ranked, i.e., expression-insensitive, features, the recognition rates approach those for neutral faces, without requiring an extensive set of reference faces for each individual to cover possible variations caused by expressions as proposed in previous work.

Ghassan Hamarneh and Xiaoxing Li, Watershed segmentation using prior shape and appearance knowledge. Journal of Image and Vision Computing, 27(1), pp. 59-68, Jan. 2009. [PDF] Abstract:Watershed transformation is a common technique for image segmentation. However, its use for automatic medical image segmentation has been limited particularly due to oversegmentation and sensitivity to noise. Employing prior shape knowledge has demonstrated robust improvements to medical image segmentation algorithms. We propose a novel method for enhancing watershed segmentation by utilizing prior shape and appearance knowledge. Our method iteratively aligns a shape histogram with the result of an improved k-means clustering algorithm of the watershed segments. Quantitative validation of magnetic resonance imaging segmentation results supports the robust nature of our method.

Xiaoxing Li and Hao Zhang, Adapting geometric attributes for expression-invariant 3D face recognition, Proceedings of Shape Modeling International (SMI), pp. 21-32, Jun. 2007. [PDF] Abstract:We investigate the use of multiple intrinsic geometric attributes, including angles, geodesic distances, and curvatures, for 3D face recognition, where each face is represented by a triangle mesh, preprocessed to possess a uniform connectivity. As invariance to facial expressions holds the key to improving recognition performance, we propose to train for the component-wise weights to be applied to each individual attribute, as well as the weights used to combine the attributes, in order to adapt to expression variations. Using the eigenface approach based on the training results and a nearest neighbor classifier, we report recognition results on the expression-rich GavabDB face database and the well-known Notre Dame FRGC 3D database. We also perform a cross validation between the two databases.

Xiaoxing Li, Greg Mori and Hao Zhang, Expression-invariant face recognition with expression classification, Proceeding of 3rd Canadian Conference on Computer and Robot Vision (CRV 06), pp. 77-84, Jun. 2006. [PDF] Abstract:Face recognition is one of the most intensively studied topics in computer vision and pattern recognition. Facial expression, which changes face geometry, usually has an adverse effect on the performance of a face recognition system. On the other hand, face geometry is a useful cue for recognition. Taking these into account, we utilize the idea of separating geometry and texture information in a face image and model the two types of information by projecting them into separate PCA spaces which are specially designed to capture the distinctive features among different individuals. Subsequently, the texture and geometry attributes are re-combined to form a classifier which is capable of recognizing faces with different expressions. Finally, by studying face geometry, we are able to determine which type of facial expression has been carried out, thus build an ex pression classifier. Numerical validations of the proposed method are given.

Xiaoxing Li and Ghassan Hamarneh, Modeling prior shape and appearance knowledge in watershed segmentation, Proceeding of 2nd Canadian conference on Computer and Robot Vision (CRV 05), pp. 21-32, May 2005 (Best Medical Imaging Paper). [PDF] Abstract:Watershed transform is widely used in image segmentation. However, its shortcomings such as over-segmentation and sensitivity to noise often make it unsuitable as an automatic tool for segmenting medical images. Utilizing prior shape knowledge has been demonstrated to improve robustness of medical image segmentation algorithms. In this paper, we propose a novel method for incorporating prior shape and appearance knowledge into watershed segmentation. Our method is based on iteratively aligning a shape-histogram with the result of an improved k-means clustering algorithm. No human interaction is needed in the whole process. We demonstrate the robustness of our method through segmenting the corpora callosa from a set of 51 brain magnetic resonance (MR) images. Numerical validation of the results is provided.

Last update March 2009.