From Terabyte-Sized Stem Cell Images to Knowledge
Peter Bajcsy, PhD
Information technology Laboratory
National Institute of Standards and Technology
10:00am Monday, 10 March 2014, ITE 346, UMBC
This talk will present the computational challenges and approaches to knowledge discovery from terabyte-sized images. The motivation comes from experimental systems for imaging and analyzing human pluripotent stem cell cultures at the spatial and temporal coverage of colonies that lead to terabyte-sized image data. The objective of such an unprecedented cell study is to characterize pluripotency of stem cell colonies over time at high statistical significance in order to understand the stem cell culture quality parameters and guide a repeatable growth of high quality stem cell colonies. The terabyte- sized images represented a stem cell line that was engineered to produce green fluorescent protein (GFP) under the influence of Oct4 promoter and then imaged in a mosaic of contiguous frames covering approximately 180 square millimeters, over five days under both phase contrast and GFP channels.
We overview multiple computer and computational science problems related to correcting (flat-field, dark current and background), stitching, segmenting, tracking, re-projecting and then representing large images for interactive visualization and sampling in a web browser. We researched extensions to Amdahl’s law for Map-Reduce computations, established benchmarks for image processing on a Hadoop platform, and introduced cluster node utilization coefficients for modeling memory demanding computations running on a computer cluster/cloud. The theoretical aspects of algorithmic complexity and cluster utilization at terabyte scale are extended to the experimental aspects of efficient image representation and client-server workload distribution in the context of visualization interactivity and image sampling. We report such experimental results for the NIST extensions to the Deep Zoom paradigm. The presentation will conclude with illustrations of enabled stem cell discoveries and collaboration opportunities to create a reference resource not only for cell biologists but also for computer scientists focusing on terabyte scale image analyses.
Peter Bajcsy received his Ph.D. in Electrical and Computer Engineering in 1997 from the University of Illinois at Urbana-Champaign and a M.S. in Electrical and Computer Engineering in 1994 from the University of Pennsylvania. He worked for machine vision, government contracting, and research and educational institutions before joining the National Institute of Standards and Technology (NIST) in 2011. At NIST, he has been leading a project focusing on the application of computational science in biological metrology, and specifically stem cell characterization at very large scales. Peter’s area of research is large-scale image-based analyses and syntheses using mathematical, statistical and computational models while leveraging computer science fields such as image processing, machine learning, computer vision, and pattern recognition. He has co-authored more than more than 24 journal papers and eight books or book chapters, and close to 100 conference papers.