Publications

back
F. Bechtold (2020)

Getting Insight on Animal Behaviour through Interactive Visualization of Multiple T-Maze Ensembles. Das Gate-O-Gon

communication medium

Bachelor's Thesis

Abstract

Behaviourists and ethologists study cognitive abilities such as learning and memory in rodents to get a better understanding of how similar processes in humans proceed. Often such studies are based on experiments of rodents placed and observed in a Multiple T-Maze. There, the path of the animals are recorded as they move inside the maze,and the resulting trajectories are then analysed. State-of-the-art analysis is based on descriptive parameters and standard statistics where one trajectory at a time is analysed. Usually it is not possible to examine multiple animal paths simultaneously. Together with experts on the field we abstracted the typical work-flow of such analyses and developed an interactive visual analytics tool, with the goal to facilitate the experts’ work andenable a deeper and novel understanding of the learning ability and decision making in rodents. After giving an overview of related works and computer-aided analysis tools in the beginning, the analysis demands and task-breakdown is presented, followed by an Explanation of the data acquisition process, data pre processing and aggregation. The underlying data structure will be explained as well. The developed analysis tool — the T-Maze Explorer — supports multiple, linked views, which support several traditional methods of visualizations, as well as two newly proposed visualizations fitted to meet the experts’ analysis demands. The first view — the T-Maze View — displays all trajectories of an ensemble with additional options such as highlighting the return path. The purpose of the second view — the Gate-O-Gon view — is to extract information from the trajectories on how often returns in the path occurred and between which parts of the maze these occurred. This information is depicted in a compact and informative novel visualization. The purpose of the T-Maze Explorer is to enable its user to easily find patterns in the data and identify irregular behaviour while inspecting a single path,multiple or the whole trajectory ensemble simultaneously. This thesis provides an insight on how the proposed visualizations were developed, the T-Maze Explorer’s characteristics and benefits as well as it’s limitations. Lastly, a brief excerpt is given on how the T-Maze Explorer could be extended in the future.

research topic

research groups

solutions