Yanhui Liang Ph.D. dissertation defense: Integrative Image and Spatial Analytics for Three-Dimensional Digital Pathology

Date of Event: 
Thursday, December 14, 2017

Location: New Computer Science 220

Time: 2-4pm

Candidate: Yanhui Liang

Program: Biomedical Informatics

Advisor: Dr. Fusheng Wang

Dissertation title: Integrative Image and Spatial Analytics for Three-Dimensional Digital Pathology


3D digital pathology provides extreme scale quantitative data with high potential for basic research in a wide scope, and becomes an emerging technology promising to support computer aided diagnosis. Quantitative analyses of 3D pathology images involve both deriving 3D micro-anatomic objects and their features, and exploring spatial relationships among a massive number of biological objects. However, this is challenged by the overwhelming data scale and 3D pathology complexity.

Our goal is to create a scalable and effective 3D digital pathology analytics framework for large-scale 3D pathology imaging data. The framework will provide novel methods on pathology image registration, segmentation, and reconstruction to transform voxel-based information from extremely large-scale 3D imaging data to 3D spatial objects. It will also provide a highly effective and scalable 3D spatial data management and querying system that enables efficient discovery of spatial patterns of 3D pathology objects.

For 3D pathology image analysis, we propose a dynamic multi-resolution approach for registration of slides from serial sections. We introduce an improved formulation with directed prior information on vessel tube probability within a variational level set framework for vessel segmentation. We propose a two-stage object cross-section association approach for 3D reconstruction with local bi-slide vessel mapping followed by a global vessel structure association via a geometrical model fitting method and Maximum A Posteriori (MAP) estimation.

For 3D spatial queries and analytics, we develop a highly scalable and efficient in-memory based 3D spatial data management and querying system. To achieve low latency, iSPEEED stores 3D objects in memory in a highly compressed form with successive levels of detail. To minimize search space and computation cost, iSPEED provides global spatial indexing in memory through partitioning at multiple levels. iSPEED provides an in-memory 3D spatial query engine, which can be invoked on-demand for running many instances in parallel. The parallelization of queries is implemented in, but not limited to, MapReduce. At run time, iSPEED dynamically decompresses only needed 3D objects at the specified level of detail, and creates necessary spatial indexes in-memory to accelerate query processing, such as on-demand inter-object-level indexing and structural indexing for individual complex structured objects such as vessels. 

Digital Minds author, Arlindo Oliveira, is talking at IACS Nov 15 !

Date of Event: 
Wednesday, November 15, 2017

Digital Minds author, Arlindo Oliveira, is talking at IACS Nov 15 !

For a preview of topic see MIT Press podcast.

Location: Institute for Advanced Computational Sciences at Stony Brook

University, IACS, Laufer Center Auditorium (map) 11:00-12:00pm

Exponential growth is a pattern built deep into the scheme of life, but technological change now promises to outstrip even evolutionary change, a process that created life and intelligence on Earth. In particular, exponential growth has characterized computing technologies in the last century, and has fueled the rapid developments of ICT that have changed society so deeply. In this talk, I will discuss the possibility that advances in computing technology will enable us to create digital minds, either artificial or natural, and discuss briefly the social, legal, and ethical implications of such a possibility.

Short​ ​bio​: Arlindo Oliveira obtained his BSc and MSc degrees in EECS from IST - Técnico Lisboa and his PhD, also in EECS, from the University of California, Berkeley. His research interests are centered in the areas of Computational Biology, Machine Learning, Computer Architecture, Algorithms and Complexity. He is the author of the book "The Digital Mind", published by MIT Press and of more than 100 articles. He became president of IST - Technical University of Lisbon - in 2012, after a career that included a number of positions in both Academy and Industry. 

Using state-space models to infer the dynamics of gene expression driven by external and internal regulatory networks

Date of Event: 
Saturday, January 14, 2017

Assistant Professor Dr. Daifeng Wang presented an invited talk at the systems genomics workshop in the PAG XXV Conference on January 14th to the 18th in San Diego, California. The title of his presentation is “Using state-space models to infer the dynamics of gene expression driven by external and internal regulatory networks”. More on the conference program  

Analyzing Open Data in Healthcare Using Public APIs and Reproducible Workflows

Date of Event: 
Friday, May 20, 2016

Dr. Janos HajagosChief of Data Analytics and Research Assistant Professor, gave a workshop, "Analyzing Open Data in Healthcare Using Public APIs and Reproducible Workflows," at the Open Data Science Conference at the Boston Exhibition and Conference Center on Friday, May 20, 2016.

Integrative Informatics and Predictive Modeling Support for Population Health

Date of Event: 
Saturday, November 14, 2015

This project targets development of proactive population health informatics methods in a geographically distributed medically and culturally heterogeneous population of app. 450,000 lives (Medicaid and Uninsured) served by 3000+ providers. The informatics infrastructure will be generally applicable to population health clinical informatics efforts. It is being developed in the context of the Suffolk Care Collaborative, a Performing Provider System (PPS) in Suffolk County, NY, funded for 5 years by the Center for Medicare and Medicaid (CMS) through the Delivery System Reform Incentive Payment (DSRIP) Program. Stony Brook University Hospital (SBUH) is the PPS Lead. Future payments to the PPS will be directly tied to performance on the key metrics of 25% reduction in potentially preventable hospitalizations (PPR) and emergency department visits (PPV) [3M Health Information Systems] and to improvements in patient quality indicators such as Agency for Healthcare Research and Quality (AHRQ) Prevention Quality Indicators (PQI) and Pediatric Quality Indicators (PDI). Eleven intervention projects (see center panel) deployed across the PPS comprise a data-generating ecosystem in which a change in one place affects outcomes of the whole. From a bioinformatics viewpoint, it is not enough to merely identify outcomes and trends; we aim to design and use an in silico test bed to evaluate and thus steer proposed interventions or changes to the system. (PDF)



Subscribe to RSS - Talks