Yanhui Liang

Ph.D. Candidate
Department of Computer Science, Department of Biomedical Informatics
Stony Brook University

NEWS: I am in the job market now!

I am looking for challenging opportunities in the domain of big data analytics, computer vision, medical imaging and machine learning. If you have any such opportunity, please let me know. Thanks!


Email: yanhui.liang at stonybrook dot edu
Address: 814 Saratoga Ave., San Jose, CA 95129
Homepage: http://bmidb.cs.stonybrook.edu/yliang/

Research Interests

Scalable Big Data Management and Analytics, Medical Imaging, Computer Vision

About me

I am a Ph.D. candidate at Department of Computer Science and Department of Biomedical Informatics at Stony Brook University, advised by Dr. Fusheng Wang. I received my M.S. degree in Computer Science from Peking University (PKU) under the supervision of Dr. Yizhou Wang in 2012, and my B.S. degree in Computer Science and Technology from Beijing University of Posts and Telecommunications (BUPT) in 2009. Prior to joining Stony Brook University, I studied at Department of Computer Science and Mathematics at Emory University.

My research interests include scalable big data management and analytics, medical imaging, computer vision and machine learning. The specialty of my research is to deal with three-dimensional (3D) data from real world, and my research projects involve large-scale 3D spatial data analysis, 3D pathology imaging and 3D scene reconstruction.


3D Spatial Big Data Analytics

iSPEED is an effective and scalable in-memory based spatial querying system for large-scale 3D data. iSPEEED stores data in memory in a highly compressed form with successive levels of detail for each individual 3D object. iSPEED provides multiple level of spatial indexing in memory through spatial data partitioning to minimize search space and computation cost. iSPEED provides an in-memory 3D spatial query engine INTENSE, which can be invoked on-demand for running many instances in parallel. The parallelization of queries is implemented in, but not limited to, MapReduce. iSPEED supports multiple spatial queries, including spatial joins, nearest neighbor search and spatial proximity estimation, and can be easily extended to others.

iSPEED provides features such as:
  • High performance spatial queries and analytics for 3D spatial big data
  • In-memory data storage with multiple Levels of Detail (LODs)
  • Multi-level 3D spatial indexing and structural indexing for complex objects
  • On-demand in-memory spatial query engine
  • 3D spatial data partitioning
  • Dynamic boundary objects handling

  • 3D Blood Vessel Reconstruction with Whole Slide Images

    As 2D microscopy images can only present 3D histopathological structures at discrete planes, they present significant information loss. To accurately capture the 3D features of micro-anatomic objects in pathology images, we propose a framework for 3D blood vessel reconstruction with 3D whole slide image volumes. The framework first performs image registration on the 3D image volume to align all images into the same coordinate space. Then we extract the boundary of blood vessels with level-set based image segmentation, and associate vessels across sections via multiple objects tracking. With associated blood vessels, we reconstruct 3D vessels by volumetric rendering for visualization.

    The framework consists of several components such as:
  • Rigid and non-rigid image registration
  • 2D objects segmentation
  • Multiple objects tracking (cross-section association)
  • 3D volumetric rendering

  • Experience

    Development of A Video Analysis Benchmark (Summer 2016, Hewlett Packard Enterprise (HPE) Labs)

    Developed and demonstrated an open-source based video analytics benchmark to evaluate the performance of edgeline systems for IoT (Internet of Things) applications at HPE.

    3D Engine View Reconstruction (Summer 2015, General Electric (GE) Global Research Center)

    Developed a pipeline to reconstruct 3D engine views from a monocular video by image matching and structure from motion. Image stitching technique is used to generate panoramic views for engine crack detection.

    2D to 3D Video Conversion (2012-2015, Peking University)

    Developed a 2D to 3D video conversion system to convert traditional 2D videos to stereoscopic ones by depth map estimation for each frame. The system has been commercialized and used by multiple third parties.


  • Yanhui Liang, Fusheng Wang, Pengyue Zhang, Joel Saltz, Daniel J. Brat, Jun Kong: Development of a Framework for Large Scale Three-Dimensional Pathology and Biomarker Imaging and Spatial Analytics. AMIA 2017

  • Pengyue Zhang, Fusheng Wang, George Teodoro, Yanhui Liang, Daniel J. Brat, Jun Kong: Automated level set segmentation of histopathologic cells with sparse shape prior support and dynamic occlusion constraint. ISBI 2017: 718-722

  • Yanhui Liang, Fusheng Wang, Darren Treanor, Derek R. Magee, Nick Roberts, George Teodoro, Yangyang Zhu, Jun Kong: A framework for 3D vessel analysis using whole slide images of liver tissue sections. I. J. Computational Biology and Drug Design 9(1/2): 102-119 (2016)

  • Yanhui Liang, Jun Kong, Fusheng Wang: Three-Dimension (3D) Blood Vessel Reconstruction and Spatial Analytics with Whole-Slide Histological Images. AMIA 2016

  • Yanhui Liang, Hoang Vo, Ablimit Aji, Jun Kong, Fusheng Wang: Scalable 3D spatial queries for analytical pathology imaging with MapReduce. SIGSPATIAL/GIS 2016: 52:1-52:4

  • Yanhui Liang, Jun Kong, Fusheng Wang: Three-dimension Whole-slide Histological Image Analytics. Pathology Informatics Summit 2016

  • Jun Kong, Pengyue Zhang, Yanhui Liang, George Teodoro, Daniel J. Brat, Fusheng Wang: Robust cell segmentation for histological images of Glioblastoma. ISBI 2016: 1041-1045

  • Hoang Vo, Jun Kong, Dejun Teng, Yanhui Liang, Ablimit Aji, George Teodoro, Fusheng Wang: Cloud-Based Whole Slide Image Analysis Using MapReduce. DMAH@VLDB 2016: 62-77

  • Yanhui Liang, Fusheng Wang, Darren Treanor, Derek R. Magee, George Teodoro, Yangyang Zhu, Jun Kong: Liver whole slide image analysis for 3D vessel reconstruction. ISBI 2015: 182-185

  • Jun Kong, Fusheng Wang, George Teodoro, Yanhui Liang, Yangyang Zhu, Carol Tucker-Burden, Daniel J. Brat: Automated cell segmentation with 3D fluorescence microscopy images. ISBI 2015: 1212-1215

  • Yanhui Liang, Fusheng Wang, Darren Treanor, Derek R. Magee, George Teodoro, Yangyang Zhu, Jun Kong: A 3D Primary Vessel Reconstruction Framework with Serial Microscopy Images. MICCAI (3) 2015: 251-259

  • Yanhui Liang, Jun Kong, Yangyang Zhu, Fusheng Wang: Three-Dimensional Data Analytics for Pathology Imaging. Big-O(Q)/DMAH@VLDB 2015: 109-125

  • Yanhui Liang, Yizhou Wang, Eric Saund: A Method of Evaluating Table Segmentation Results Based on a Table Image Ground Truther. ICDAR 2011: 247-251