Yi Zhen

Ph.D Candidate
Department of Software and Information Systems ,
College of Computing and Informatics,
University of North Carolina at Charlotte


Yi Zhen is a PhD student in Department of Software and Information Systems at University of North Carolina at Charlotte under the supervision of Prof. Yaorong Ge
Her current research interest is focusing on Health Informatics, Machine Learning, Computer Aided Diagnosis, Medical Image Processing and Analysis, Medical Imaging.
(Curriculum Vitae).

Research Experience:

Research assistant in Biomedical Engineering Department of Wake Forest University, my research is focusing on development of an imaging informatics and analytics platform for cardiovascular research.

Research assistant in Center for Biomedical Imaging and Bioinformatics (CBIB) , my  research included image segmentation, medical image analysis, optical image restoration, 3D visualization, hand detection and recognition.


  • PhD. candidate of Department of Software and Information Systems, University of North Carolina at Charlotte, U.S.A  2013-Present  
  • PhD. candidate of Biomedical Engineering Department, Wake Forest University, U.S.A  2012-2013 
  • M.S.,  Biomedical Engineering, Huazhong University of Science and Technology, China  2010-2012  
  • B.S.,  Software Engineering, Huazhong University of Science and Technology, China  2006-2010  

Projects & Publications:

Novel Supervised C-V Segmentation
A novel hybrid segmentation algorithm was proposed based on Otsu method and C-V model. Not like the common hybrid segmentation methods, a new penalty function was defined to measure the difference between the level set function and the results segmented by Otsu method. It also provided a new framework for combining active contours with other segmentation methods. In the new framework to combine Otsu method and C-V model, results segmented by Otsu method was used to restrain the deforming of the level set function in C-V model.

3-D Segmentation of Pulmonary Nodule
This project aimed to develop a Lung CT based Pulmonary Nodules Computer-Aided Detection CADsystem for detecting nodules and extracting pathologic features. A 3D segmentation method based on global minimum active contour was proposed to effectively segment nodules. The novel method utilized both of edge and region information to segment different types of nodules. Besides, the CAD system provided features extraction and nodule 3D visualization.

3-D Reconstruction + VTK

The reconstruction and deformation was based on Visualization Toolkit 5.0 (VTK). Two method in three dimensional image reconstruction were compared, one is surface rendering and the other is volume rendering. Besides, the performance of three algorithms were compared on volume rendering, including ray casting, splatting and shear-warp. And the reconstruction algorithms were implemented by extending a VTK class to reconstruct and deform the surface of pulmonary nodules. The source code is written in C++ with OpenCV and VTK.



This project aimed to implement gesture motion recognition system, consisting of three modules: image collecting, hand segmentation, and gesture recognition. Currently, hand segmentation modules has been completed by using skin color segmentation. Firstly, dynamic background modeling was based on two different models, one is weighted average model and the other is Gaussian mixture model. Secondly, a threshold was imported using face color as hand’s color cluster threshold. The face recognition was implemented by using Haar feature recognition method.

Intelligent Diagnosis Report System


The project aimed to develop a diagnosis report system of lung CT CAD system. The system is written in C# language under .NET framework. The main function is generating lung CT report automatically.  Specifically, it contains translating computational diagnosis data into natural language within a structured format, then filling the diagnosis data automatically, and supporting “search suggest” in diagnostic terms. The diagnosis report system is welcomed in hospitals. 


Operating Rooms Scheduling and Planning System


The project was my undergraduate thesis. The innovative point is importing an ORs scheduling model and a new penalty function based on simulated annealing algorithm. The system was implemented in C# and tested its performance with data from hospitals. 







Packing Unequal Circles 


The project belonged to course project of Algorithms Analysis. It is the implementation of an energy strategy for packing unequal circles into a circle container. 


Contact Me:

Address:Woodward Hall 350, 
              Department of Software and Information Systems ,
              College of Computing and Informatics,
              University of North Carolina at Charlotte
Email: yzhen[at]uncc[dot]edu