Paper Info
Title
Three-Dimensional Data Analytics for Pathology Imaging.
Abstract
Three-dimensional (3D) structural changes and spatial relationships of micro-anatomic objects in whole-slide digital pathology images encode a large wealth of information on normal tissue development and disease progression. In this paper, we present a complete framework for quantitative spatial analytics of 3D micro-anatomic objects with pathology image volumes, with special focus on vessels and nuclei. Reconstructing 3D vessel structures from a sequence of whole-slide images, we simulate 3D biological systems by generating 3D nuclei uniformly distributed around 3D vessels. Given nuclei are distributed around vessels, intersection detection with 3D nuclei and vessels is conducted by a heuristic algorithm with data structure Axis-Aligned Bounding-Box (AABB). Motivated by real-world use case, we also travel the AABB tree constructed from 3D vessel structures to perform the distance-based query between 3D nuclei and vessels. We quantitatively evaluate the performance of 3D intersection detection by the heuristic algorithm and distance-based query based on AABB tree traversal. Experimental results demonstrate the efficiency of our framework for 3D spatial analytics with whole-slide serial pathology image dataset.
Year
DOI
Venue
2015
10.1007/978-3-319-41576-5_8
BIOMEDICAL DATA MANAGEMENT AND GRAPH ONLINE QUERYING
Keywords
Field
DocType
Pathology imaging,3D data simulation,Spatial analytics,Distance-base query,AABB Tree
Spatial analysis,ENCODE,Data mining,Data structure,Tree traversal,Data analysis,Heuristic (computer science),Computer science,Disease progression,Digital pathology,Pathology
Conference
Volume
ISSN
Citations 
9579
0302-9743
0
PageRank 
References 
Authors
0.34
12
4
Name
Order
Citations
PageRank
Yanhui Liang1166.50
Jun Kong223729.70
Yangyang Zhu371.24
Fusheng Wang4102679.28