Home Research Computer Graphics Research Medical Image Analysis

Medical Image Analysis

Deformable Segmentation using Sparse Shape Representation

Organ shape plays an important role in various clinical practices such as segmentation. Effective modeling of shape priors is challenging because: (1) shape variation is complex and cannot always be modeled by a parametric probability distribution; (2) a shape instance derived from image appearance cues (input shape) may have gross errors; and (3) local details of the input shape are difficult to preserve if they are not statistically significant in the training data. We propose Sparse Shape Composition model (SSC) to deal with these three challenges in a unified framework. In our method, a sparse set of shapes in the shape repository is selected and composed together to infer/refine an input shape. The a priori information is thus implicitly incorporated on-the-fly. It is formulated as a sparse learning problem, and is extensively validated on several medical applications, including 2D lung localization in X-ray images and 3D liver segmentation in low-dose CT scans.

Project page and code: http://www.research.rutgers.edu/~shaoting/research/siemens2010/project.htm

Publications

  • Deformable Segmentation via Sparse Shape Representation. Shaoting Zhang, Yiqiang Zhan, Maneesh Dewan, Junzhou Huang, Dimitris Metaxas and Xiang Zhou. MICCAI, The 14th Annual International Conference on Medical Image Computing and Computer Assisted Intervention, 2011. MICCAI Young Scientist Award Finalist
  • Towards Robust and Effective Shape Modeling: Sparse Shape Composition. Shaoting Zhang, Yiqiang Zhan, Maneesh Dewan, Junzhou Huang, Dimitris Metaxas and Xiang Zhou. MedIA, Medical Image Analysis, Volume 16, Issue 1, Pages 265-277, January 2012.


 


 

    Bioluminescence Image Anaysis and Diagnosis

    This work introduces a novel and efficient algorithm for reconstructing the 3D shapes of tumors from a set of 2D bioluminescence images which are taken by the same camera but after continually rotating the animal by a small angle. The method is efficient and robust enough to be used for analyzing the repeated imaging of a same animal transplanted with gene marked cells. There are several steps in our algorithm. First, the silhouettes (or boundaries) of the animal and its interior hot spots (corresponding to tumors) are segmented in the set of bioluminescence images. Second, the images are registered according to the projection of the animal rotating axis. Third, the images are mapped onto 3D projection planes and from the viewpoint of each plane, the visual hulls of the animal and its interior tumors are reconstructed. Then, the intersection of visual hulls from all viewpoints approximates the shape of the animal and its interior tumors. In order to visualize in 3D the structure of the tumor, we also co-register the BLI-reconstructed crude structure with detailed anatomical structure extracted from high-resolution micro-CT on a single platform. The experimental results show promising performance of our reconstruction and co-registration method.

    Publications

    • Junzhou Huang, Xiaolei Huang, Dimitris Metaxas, Debarata Banerjee, ”3D Tumor Shape Reconstruction from 2D Bioluminescence Images and Registration with CT Images”, 1st Workshop on Microscopic Image Analysis with Applications in Biology, MIAAB’06, 2006. (Oral)
    • Junzhou Huang, Xiaolei Huang, Dimitris Metaxas, Debarata Banerjee, ”3D Tumor Shape Reconstruction from 2D Bioluminescence Images” , IEEE Int’l Symposium on Biomedical Imaging: From Nano to Macro, ISBI’06, pp. 606-609, 2006. (Oral)

    Project Homepage

    4D High-resolution Cardiac Reconstruction

    Recent developments on 320 multi-detector CT technologies have made the volumetric acquisition of 4D high resolution cardiac images in a single heart beat possible. In this paper, we present a framework that uses these data to reconstruct the 4D motion of the endocardial surface of the left ventricle for a full cardiac cycle. This reconstruction framework captures the motion of the full 3D surfaces of the complex anatomical features, such as the papillary muscles and the ventricular trabeculae, for the first time, which allows us to quantitatively investigate their possible functional significance in health and disease.

    Publications

    • Mingchen Gao, Junzhou Huang, Shaoting Zhang, Zhen Qian, Szilard Voros, Dimitris Metaxas, Leon Axel, "4D Cardiac Reconstruction Using High Resolution CT Images", FIMH2011

    Project Homepage

     


     

      Sparse Regularization for Tag Separation in tMRI

      We introduce a tag separation method for better cardiac boundary segmentation and tag tracking. Our approach is based on two observations in the cardiac tagged MR images: 1) the tag patterns have a regular texture; 2) the cardiac images without tag patterns are piecewise smooth with sparse gradients. These observations motivate us to use two dictionaries, one based on the Discrete Cosine Transform for representing tag patterns and the other based on the Wavelet Transform for representing the underlying cardiac image without tag patterns. The two dictionaries are built such that they can lead to sparse representations of the tag patterns and of the piece-wise smooth regions without tag patterns. With the two dictionaries, a new tag separation approach is proposed to simultaneously optimize w.r.t. the two sparse representations,where optimization is directed by the Total Variation regularization scheme. While previous methods have focused on tag removal,our approach to acquiring both optimally-decomposed tag-only image and the cardiac image without tags simultaneously can be used for bettertag tracking and cardiac boundary segmentation. We demonstrate the superior performance of the proposed approach through extensive experiments on large sets of cardiac tagged MR images.

      Publications

      • Junzhou Huang, Zhen Qian, Xiaolei Huang, Dimitris Metaxas, Leon Axel, ”Tag Separation in Cardiac Tagged MRI”, In Proc. of the 11th Annual International Conf. on Medical Image Computing and Computer Assisted Intervention, MICCAI’08, LNCS-5242, pp. 289-297, 2008.

      Project Homepage

      4D Cardiac Reconstruction and Blood Flow Simulation

      Following a heart attack or the development of some cardiovascular diseases, the movement of the heart walls during the cardiac cycle may change, which affects the motion of blood through the heart, potentially leading to an increased risk of thrombus. While Doppler ultrasound and MRI can be used to monitor valvular blood flow, the image resolutions are low and they cannot capture the interactions between the highly complex heart wall and the blood flow. For this reason, with the rapid development of high-resolution cardiac CT, patient-specific blood flow simulation can provide a useful tool for the study of cardiac blood flow.

      Recent developments on CT technologies have made the acquisition of high-resolution 4D  cardiac images possible. We present a framework that uses this data to reconstruct the 4D motion of the endocardial surface of the left ventricle for a full cardiac cycle. This reconstruction framework captures the motion of the full 3D surfaces of the complex anatomical features, such as the papillary muscles and the ventricular trabeculae, for the first time.

      We then present a method to simulate and visualize blood flow through the heart, using the reconstructed 4D motion of the endocardial surface of the left ventricle as boundary conditions. The reconstruction captures the motion of the full 3D surfaces of the complex features, such as the papillary muscles and the ventricular trabeculae. We use visualizations of the flow field to view the interactions between the blood and the trabeculae in far more detail than has been achieved previously, which promises to give a better understanding of cardiac flow.

      Publications

      • Scott Kulp, Mingchen Gao, Shaoting Zhang, Zhen Qian, Szilard Voros, Leon Axel, and Dimitris Metaxas. "Using High Resolution Cardiac CT Data to Model and Visualize Patient-Specific Interactions Between Trabeculae and Blood Flow." Proceedings of the Fourteenth International Conference on Medical Image Computing and Computer Assisted Intervention. 2011.
      • Mingchen Gao, Junzhou Huang, Shaoting Zhang, Zhen Qian, Szilard Voros, Dimitris Metaxas, and Leon Axel. "4D Cardiac Reconstruction Using High Resolution CT Images." Proceedings of the Sixth International Conference on Functional Imaging and Modeling of the Heart. 2011.
      • Scott Kulp, Dimitris Metaxas, Zhen Qian, Szilard Voros, Leon Axel, and Viorel Mihalef. "Patient-Specific Modeling and Visualization of Blood Flow through the Heart." Proceedings of the Eighth International Symposium on Biomedical Imaging. 2011.
      • Viorel Mihalef, Dimitris Metaxas, Mark Sussman, Vassilios Hurmusiadis, Leon Axel. "Atrioventricular blood flow simulation based on patient-specific data". Proceedings of FIMH 2009.