SPX: A Versatile Spatial Indexing Framework

Information

• Publication Type: Master Thesis
• Workgroup(s)/Project(s): not specified
• Date: 2024
• Open Access: yes
• First Supervisor: Eduard Gröller
• Pages: 157
• Keywords: spatial indexing, volume data, brain atlas

Abstract

The integration of modern genetic techniques, advanced volumetric imaging methods, and single-cell extraction methods has empowered neurobiologists to create extensive digitized and coregistered specimen sample collections. These collections serve as valuable resources for studying neuronal structures, functional compartments, and neurological connections within the brain. By sampling single cells from various locations within the animal brain, scientists can investigate cell type distributions and gene expressions. However, the exploration of these vast collections, which include volumetric images, segmented structures, and gene expression data, poses a significant challenge in neuroscience. Efficient access to specific regions of interest in all images, derivative processed data, cell samples, and metadata is essential for researchers. In this thesis, I present a flexible and extensible approach to spatially index and store volumetric grid data and region-based data, enabling efficient access and providing a streamlined method for implementing new data abstractions, query types, and indexing strategies. The framework supports different datasets, the encoding of neurological structural types, and incorporates a layering mechanism to handle multiple data representations or time-dependent data within a single data structure. Standardized interfaces are defined for loading voxel and region data, preprocessing them, creating data abstractions, and implementing new query types. The data storage is managed using a storage engine approach, allowing users to leverage different storage mechanisms or introduce their own.This thesis provides an overview of conceptual ideas, implementation details, current data abstractions, and query types. The system was evaluated in terms of performance and scalability in its current use cases. A short introduction to three applications, BrainBaseWeb, BrainTrawler, and BrainTrawler Lite, exemplifies the usage of this framework.

Additional Files and Images

Weblinks

• Entry in reposiTUm (TU Wien Publication Database)
• DOI: 10.34726/hss.2024.119688

BibTeX

@mastersthesis{toepfer-2024-spx,
  title =      "SPX: A Versatile Spatial Indexing Framework",
  author =     "Markus T\"{o}pfer",
  year =       "2024",
  abstract =   "The integration of modern genetic techniques, advanced
               volumetric imaging methods, and single-cell extraction
               methods has empowered neurobiologists to create extensive
               digitized and coregistered specimen sample collections.
               These collections serve as valuable resources for studying
               neuronal structures, functional compartments, and
               neurological connections within the brain. By sampling
               single cells from various locations within the animal brain,
               scientists can investigate cell type distributions and gene
               expressions. However, the exploration of these vast
               collections, which include volumetric images, segmented
               structures, and gene expression data, poses a significant
               challenge in neuroscience. Efficient access to specific
               regions of interest in all images, derivative processed
               data, cell samples, and metadata is essential for
               researchers. In this thesis, I present a flexible and
               extensible approach to spatially index and store volumetric
               grid data and region-based data, enabling efficient access
               and providing a streamlined method for implementing new data
               abstractions, query types, and indexing strategies. The
               framework supports different datasets, the encoding of
               neurological structural types, and incorporates a layering
               mechanism to handle multiple data representations or
               time-dependent data within a single data structure.
               Standardized interfaces are defined for loading voxel and
               region data, preprocessing them, creating data abstractions,
               and implementing new query types. The data storage is
               managed using a storage engine approach, allowing users to
               leverage different storage mechanisms or introduce their
               own.This thesis provides an overview of conceptual ideas,
               implementation details, current data abstractions, and query
               types. The system was evaluated in terms of performance and
               scalability in its current use cases. A short introduction
               to three applications, BrainBaseWeb, BrainTrawler, and
               BrainTrawler Lite, exemplifies the usage of this framework.",
  pages =      "157",
  address =    "Favoritenstrasse 9-11/E193-02, A-1040 Vienna, Austria",
  school =     "Research Unit of Computer Graphics, Institute of Visual
               Computing and Human-Centered Technology, Faculty of
               Informatics, TU Wien",
  keywords =   "spatial indexing, volume data, brain atlas",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2024/toepfer-2024-spx/",
}