[Home] [Education] [Working Experience] [Projects] [TA Work] [Publications] [Courses]


Zhennan Yan is currently a Ph.D. candidate working in Computer Science at Rutgers, the State University of New Jersey. His advisor is Prof. Dimitris N. Metaxas, and co-advised by Shaoting Zhang.

His research interests are medical image analysis, deformable models, machine learning methods.

(Curriculum Vitae).


Education
  • 2010 - present, graduate student in Computer Science, Rutgers University
  • 2006 - 2009, M.S. in Computer Applied Technology, Shanghai Jiao Tong University, China
  • 2002 - 2006, B.E. in Software Engineering, Tongji University, China

  • Working Experience
  • 06/2015 - 08/2015, Scientist Intern at Siemens Syngo R&D US group
  • 06/2014 - 08/2014, Scientist Intern at Siemens Syngo R&D US group
  • 06/2013 - 08/2013, Intern Consultant at BioClinica Inc.
  • 07/2012 - present, Research Assistant at Rutgers University
  • 09/2010 - 06/2012, Teaching Assistant at Rutgers University
  • 2009 - 2010, Software Engineer in Shanghai, China
  • 01/2007 - 03/2009, Research Assistant at Shanghai Jiao Tong University
  • 05/2008 - 09/2008, Intern Software Develop Engineer, SSG-IPAT group, Intel Asia-Pacific R&D Ltd.
  • Zhennan Yan (闫桢楠)
    Department of Computer Science Rutgers University
    Piscataway, NJ 08854

    Resume ►
    Projects
     

    Supervised discriminative patch discovery (Bodypart recognition)

    This is an project for human Body part recognition in transversal CT images.
     Automatic medical image analysis systems need human body part identification in the image.
     Slice-based bodypart recognition is the key of 3D bodypart identification: given a transversal slice, recognize which body part it comes from; image classification problem.
     Characteristic: the body part of a slice is usually identified by local discriminative regions instead of global image context, e.g., a cardiac slice is differentiated from an aorta arch slice by the mediastinum region.
     Two key questions: (1) Which local regions are discriminative for bodypart recognition? (2) How to learn the local bodypart identifiers on them without time-consuming manual annotations?
     Solution: Multi-stage deep learning to “discover” the most discriminative local patches and learn classifier afterwards.
    Intern work at Siemens Healthcare, Malvern, PA, USA.

     

    Human brain segmentation

    This is an on-going project for human brain parcellation in structural MR images. The project is designed to segment human brain into 30+ structures (including cortical GM, WM, CSF and sub-cortical structures) simultaneously and accurately. Proposed an Extended Adaptive Statistical Atlas (EASA) based approach, and a hybrid extension.
    Supported by Dr. Albert Montillo, GE Global Research, Niskayuna, NY, USA.

    • [ISBI'13] Zhennan Yan, Shaoting Zhang, Xiaofeng Liu, Dimitris N. Metaxas, Albert Montillo, AIBL research group: Accurate Segmentation of Brain Images into 34 Structures Combining A Non-Stationary Adaptive Statistical Atlas and A Multi-atlas with Applications to Alzheimer's Disease.
     

    Lung segmentation in CXR

    This is a course project for human lung segmentation in chest X-Ray images, based on landmark detection and sparse shape representation. The project is designed to make significant improvements for organ segmentation in accuracy and time cost. The application is now limited to 2D medical images. And it will be extended to 3D images later.

     

    Prostate elasticity reconstruction

    This is an on-going project for prostate elasticity reconstruction based on ultrasonic displacement and strain images. The project is designed to make significant improvements in detection and evaluation of prostate cancer by developing advanced ultrasonic elastographic methods for sensing and imaging cancerous foci in the prostate based on their relative stiffness. The project is now limited to 2D finite element model of prostrate, which segmented from ultrasound image. And it will be extended to 3D model later.
    Supported by Dr. S. Kaisar Alam, Riverside Research Institute, Lizzi Center for Biomedical Engineering

     

    Craniofacial Surgery Simulation

    This project is designed to pursue a realistic 3D simulation of craniofacial plastic surgery. The project aims to provide an easy-to-use 3D platform to surgeons and assist them to do planning before the practical surgery. It combines, extends some methods from image processing, geometric modeling and finite element method (FEM). The input is CT data of head. Appling image processing and our improved meshing algorithm, we can use our improved FEM procedure to predicate the result facial shape after truncating the prominent mandible angle.
    IGST-Lab, Shanghai Jiao Tong University.

     

    Deformation Toolkit (DTK) for Virtual Surgery Simulation

    This project aimed to design and implement an improved toolkit suitable for modeling and simulation of soft tissues, which can be easily reused in different surgery related applications. DTK is based on Finite Element Method (FEM). And the linear/nonlinear FEM computation is based on the 3D tetrahedron mesh. DTK can generate very nice tetrahedron mesh by our improved meshing algorithm which is based on Delaunay criterion and Centroidal Voronoi Tessellations structure. DTK also uses sparse structure, CUDA technology, and etc. to improve the efficiency of FEM.
    IGST-Lab, Shanghai Jiao Tong University. Team members: Zhennan Yan, Sizhe Lv.


    TA Work
      No Image

    CS323, Numerical Analysis and Computing

    TA for this course in Spring 2012. The instructor is Prof. Gerard Richter. The main aim of CS323 is to expose the student to the development, application, and analysis of basic numerical algorithms. Contents include:

    • Solution of non-linear equations; Solution of linear algebraic systems; Approximation, interpolation; Numerical differentiation and integration; Solution of ordinary differential equations.

      No Image

    CS314, Principles of Programming Languages

    TA for this course in Fall 2011. The instructor is Prof. Louis Steinberg. The main aim of CS314 is to teach students new ways to think about problems and programs. Contents include:

    • Topics of grammars (RE, NFA, DFA), parameter passing modes, types and type checking;
    • Functional programming (Scheme), logic programming (Prolog), scripting (Python).

      No Image

    CS111, Introduction to Computer Science

    TA for this course in Spring 2011. The instructor is Pradip Hari. The aim of CS111 is to introduce the student to the fundamental techniques used in computer science and software development. Upon completion of this course, the successful student should be able to design, write, test, and analyze programs to solve simple real-world problems. CS111 uses Java programming language.

      No Image

    CS110, Introduction to Computers and Applications

    TA for this course in Fall 2010. The instructors were Alan Belowich, Kate Goelz and Jt. The course gave students hands-on experience with applications in presentation software, electronic spreadsheet, database management, network applications, web page design and TRUE BASIC programming.