Master Thesis

FZJ-2022-04854

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png Parallelization and optimization of measures in Network Neuroscience

Kirchner, T. (Corresponding author)FZJ*

2022

Please use a persistent id in citations: http://hdl.handle.net/2128/32684

Abstract: The main goal of neuroscience is to understand the human brain. To achieve this, the area Network Neuroscience has evolved. This scientific field describes the human brain as a network comprising of neural elements (nodes), and connection between these elements (edges). Graph analysis algorithms can therefore be applied to gain insights into the brain. The Brain Connectivity Toolbox (BCT) is a Matlab toolbox that includes graph algorithms calculating measures in Network Neuroscience. This toolbox only provides sequential versions of these algorithms, which limits the size of the input graph. To overcome this limit, we implement optimized shared and distributed memory algorithms from this toolbox. For the implementations, the Boost Graph Library is used. This library provides a generic interface for the graph data structure. Experiments that are run on the Jülich supercomputer JUWELS show that a performance improvement can be achieved by this parallelization.

Note: Masterarbeit, FH Aachen Campus Jülich, 2022

---

Contributing Institute(s):

1. Jülich Supercomputing Center (JSC)

Research Program(s):

1. 5111 - Domain-Specific Simulation & Data Life Cycle Labs (SDLs) and Research Groups (POF4-511) (POF4-511)
2. SLNS - SimLab Neuroscience (Helmholtz-SLNS) (Helmholtz-SLNS)

Appears in the scientific report 2022

---

Database coverage:

---