Home
Up
Publications
Teaching
Links

 
  This material is based upon work supported by the National Science Foundation (Principal Investigator: Jing Hua). Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF).

 

Visual Analytics of Cross-subject, Multi-measure, Multi-dimensional Imaging Data

Jing Hua

Graphics and Imaging Lab

Wayne State University

 

Summary


Scientific imaging data is typically collected from individual subjects, and many consists of sophisticated three-dimensional (3D) geometric structures and high-dimensional, heterogeneous features. Assessment of similarity and disparity from the multimodality, heterogeneous 3D imaging data across subjects plays a central role in modern scientific discovery. In many scenarios, intrinsic geometric structures embedded in 3D imaging of real-world objects are very effective in mapping individual objects for interpretation of their similarity and disparity. A rigorous computational framework that tightly unifies geometric mapping, diffusion and matching is essential for integrative analysis of a variety of underlying relationships in features inherited from imaging datasets of a large number of subjects, i.e., 3D imaging informatics. These research activities will address the following major themes and objectives: (1) To design novel geometric diffusion theory and methods to compute multi-scale representations for imaging data abstraction and then investigate hierarchical image feature extraction and matching/registration; (2) To design and develop a coordinated visualization framework for comparative analysis of cross-subject multi-measure brain images, which includes both data abstraction and visual abstraction; (3) To validate and apply the proposed theoretical approaches to real medical imaging informatics problems, such as Brain Tumor, Epilepsy, Sturge-Weber Syndrome, and so on. In addition, the PIs' research endeavors have been tightly integrated with a complementary set of educational objectives, including: (1) the development of new strategies for truly multi-disciplinary science education; (2) the enhancement of the existing curricula; (3) the doctoral training of graduate researchers; (4) the implementation of mentoring activities for students from underrepresented groups.

 

 

 

Current Results:

 

Intrinsic Geometric Scale Space (IGSS)

 

                              .

We have formalized a novel, intrinsic geometric scale space (IGSS) of 3D surface shapes. The intrinsic geometry of a surface is diffused by means of the Ricci flow for the generation of a geometric scale space. We rigorously prove that this multiscale shape representation satisfies the axiomatic causality property. Within the theoretical framework, we further present a feature-based multiscale shape representation derived from IGSS processing, which is shown to be theoretically plausible and practically effective. By integrating the concept of scale-dependent saliency into the shape description, this representation is not only highly descriptive of the local structures, but also exhibits several desired characteristics of global shape representations, such as being compact, robust to noise and computationally efficient. We demonstrate the capabilities of our approach through salient geometric feature detection and highly discriminative matching of 3D scans.

 

 

■ Authalic Parameterization of General Surfaces Using Lie Advection

 

Parameterization of complex surfaces constitutes a major means of visualizing highly convoluted geometric structures as well as other properties associated with the surface. It also enables users with the ability to navigate, orient, and focus on regions of interest within a global view and overcome the occlusions to inner concavities. We propose a novel area-preserving surface parameterization method which is rigorous in theory, moderate in computation, yet easily extendable to surfaces of non-disc and closed-boundary topologies. Starting from the distortion induced by an initial parameterization, an area restoring diffeomorphic flow is constructed as a Lie advection of differential 2-forms along the manifold, which yields equality of the area elements between the domain and the original surface at its final state. Existence and uniqueness of result are assured through an analytical derivation. Based upon a triangulated surface representation, we also present an efficient algorithm in line with discrete differential modeling. As an exemplar application, the utilization of this method for the effective visualization of brain cortical imaging modalities is presented. Compared with conformal methods, our method can reveal more subtle surface patterns in a quantitative manner. It, therefore, provides a competitive alternative to the existing parameterization techniques for better surface-based analysis in various scenarios.

 

■ Quansi-Conformal Surface Deformation Analysis

 

      

Quasi-conformal deformation analysis of 3D dynamic models utilizes a new metric on the quotient space of surfaces which is able to capture changes of the curvature at each vertex of a simplicial complex during deformation. Hence, the deformation curve can be obtained by calculating the geodesic curve connecting two shapes in the shape space manifold. To facilitate the deformation analysis, the deformation curves are first transferred to the same location of the shape space. Finally, the Multi-Dimensional Scaling method is employed to eliminate the redundant dimensions allowing easy comparison of the deformations. The algorithm can effectively classify normal and abnormal deformations in shape space.

 

Surface Registration with Shape Spectrum

 

        We develop a fully automatic geometric method for registering  surfaces extracted from images. The algorithm uses shape spectrum to extract the shape characteristics which are employed as the surface signature to find the correspondent regions between the surfaces. The method is simple yet efficient and accurate. The novel surface registration method based on shape spectrum is of excellent accuracy for surface registration. Based on our observation, the second eigenvector of the Laplacian, called the Fiedler vector, has interesting properties, making it a good permutation vector for numerical computations.
 

Image Processing for Prepless Virtual Colonography

 

                         

A novel segmentation framework for a prepless virtual colonoscopy (VC) is presented, which reduces the necessity for colon cleansing before the CT scan. The patient is injected rectally with a water-soluble iodinated contrast medium using manual insufflators and a small rectal catheter. Compared to the air-based contrast medium, this technique can better preserve the color lumen and reduce the partial volume effect. However, the contrast medium, together with the fecal materials and air, makes colon wall segmentation challenging.Our solution makes no assumptions about the shape, size, and location of the fecalmaterial in the colon. This generality allows us to label the fecal material accurately and extract the colon wall reliably. The accuracy of our technique has been verified on 60 human subjects. Compared with current VC technologies, our method is shown to be better in terms of both sensitivity and specificity. Further, in our experiments, the accuracy of the technique was comparable to that of optical colonoscopy results.

 

■ Business Model-based Coclustering

 

               

We present a quantitative analysis and modeling tool that is able to characterize anatomical connectivity patterns based on a newly developed coclustering algorithm, termed the Business model-based Coclustering Algorithm (BCA). We apply BCA to Diffusion Tensor Imaging (DTI) data in order to provide an automated and reproducible assessment of the connectivity patterns between different cortical areas in human brains. BCA not only partitions the cortical mantel into well-defined clusters, but also maximizes the connectivity strength between these clusters. Moreover, BCA is computationally robust and allows both outlier detection as well as parameter-independent determination of the number of clusters. Our coclustering results have showed good performance of BCA in identifying major white matter fiber bundles in human brains and facilitated the detection of abnormal connectivity patterns in patients suffering from various neurological diseases.
 

■ Ricci Flow-based Spherical Parameterization and Surface Registration

 

We present an improved Euclidean Ricci ow method for spherical parameterization, and subsequently invent a scale space processing built upon Ricci energy to extract robust surface features for accurate surface registration. Since our method is based on the proposed Euclidean Ricci ow, it inherits the properties of Ricci ow such as conformality, robustness and intrinsicalness, facilitating e cient and e ective surface mapping. Compared with other surface registration methods using curvature or sulci pattern, our method demonstrates a signi cant improvement for surface registration. In addition, Ricci energy can capture local di erences for surface analysis as shown in the experiments and applications.

 

 

Software for Download:

 

Please register and download the software: BrainSpace V2.1 at the following link: http://141.217.205.25:8080/ClientApp/clientdata. The brief introduction and manual can be download at BrainSpace_V2.1.pdf.

 

Selected, Related Publications:

  1. Jiaxi Hu, Guangyu Zou, and Jing Hua, "Volume-Preserving Mapping and Registration for Collective Data Visualization," IEEE Transactions on Visualization and Computer Graphics (IEEE VIS '14), Vol. 20, No. 6, pp. 2664-2673, 2014.

  2. Vahid Taimouri and Jing Hua, "Deformation Similarity Measurement in Quasi-Conformal Shape Space," Graphical Models, Vol. 76, No. 2, pp. 57-69, 2014.

  3. Xuejiao Chen, Jiaxi Hu, Huiguang He, and Jing Hua, "Spherical Volume-Preserving Demons Registration," Computer-Aided Design (ACM SPM '14), 2014.

  4. Vahid Taimouri and Jing Hua, "Visualization of Shape Motions in Shape Space," IEEE Transactions on Visualization and Computer Graphics (IEEE VIS '13), Vol. 19, No. 12, 2644-2652, 2013.

  5. Xuejiao Chen, Huiguang He, Guangyu Zou, Xiaopeng Zhang, Xianfeng Gu, and Jing Hua, "Ricci Flow-based Spherical Parameterization and Surface Registration," Computer Vision and Image Understanding, Vol. 117, No. 9, pp. 1107-1118, 2013.

  6. Guangyu Zou, Jiaxi Hu, Xianfeng Gu, and Jing Hua, "Authalic Parameterization of General Surfaces Using Lie Advection," IEEE Transactions on Visualization and Computer Graphics (VIS), Vol. 17, No. 6, 2011.

  7. Vahid Taimouri, Xin Liu, Zhaoqiang Lai, Chang Liu, Darshan Pai, and Jing Hua, "Colon Segmentation for Prepless Virtual Colonoscopy," IEEE Transactions on Information Technology in Biomedicine, Vol. 15, 2011.

  8. Darshan Pai, Hamid Soltanian-Zadeh, and Jing Hua, "Evaluation of Fiber Bundles across Subjects through Brain Mapping and Registration of Diffusion Tensor Data," NeuroImage, Vol. 54, pp. S165-S175, 2011.

  9. Cui Lin, Darshan Pai, Shiyong Lu, Otto Muzik, and Jing Hua, Coclustering for Cross-subject Fiber Tract Analysis through Diffusion Tensor Imaging, IEEE Transactions on Information Technology in Biomedicine, Vol. 14, No. 2, pp. 514 - 525, 2010.

  10. Guangyu Zou, Jing Hua, Zhaoqiang Lai, Xianfeng Gu, and Ming Dong, "Intrinsic Geometric Scale Space by Shape Diffusion," IEEE Transactions on Visualization and Computer Graphics (VIS), Vol. 15, No. 6, pp. 1193 - 1200, 2009.

  11. Yanhua Chen, Lijun Wang, Ming Dong, and Jing Hua, "Exemplar-based Visualization of Large Document Corpus," IEEE Transactions on Visualization and Computer Graphics (INFOVIS), Vol. 15, No. 6, pp. 1169 - 1176, 2009.

  12. Guangyu Zou, Jiaxi Hu, Xianfeng Gu, and Jing Hua, "Area-preserving Surface Flattening Using Lie Advection," In Proceedings of the 14th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2011.

  13. Zhaoqiang Lai, Jiaxi Hu, Chang Liu, Vahid Taimouri, Darshan Pai, Jiong Zhu, Jianrong Xu, and Jing Hua, "Intra-patient Supine-Prone Colon Registration in CT Colonography Using Shape Spectrum," In Proceedings of the 13th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2010 (Oral; Acceptance Rate: 5.7%)

  14. Vahid Taimouri, Huiguang He, and Jing Hua, "Comparative Analysis of Quasi-Conformal Deformations in Shape Space," In Proceedings of the 13th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2010.

  15. Wei Zeng, Lok Ming Lui, Lin Shi, Defeng Wang, Winnie C.W. Chu, Jack C.Y. Cheng, Jing Hua, Shing-Tung Yau, and Xianfeng Gu, "Shape Analysis of Vestibular Systems in Adolescent Idiopathic Scoliosis Using Geodesic Spectra," In Proceedings of the 13th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2010.

 

Point of Contact: PI Jing Hua ()

Date of Last Update: June 10, 2014.