Work

A comparison study of Kdtree, Vptree and Octree for storing neuronal morphology data with respect to performance

Year: 2016
Type: dissertation
Abstract: In this thesis we investigated performance of Kdtree, Vptree and Octree for storing neuronal morphology data. Two naive list structures were implemented to compare with the space partition data structures. The performance was measured with different sizes of neuronal networks and different types of test cases. A comparison with focus on cache misses, average search time and memory usage was made. Furthermore, measurements gathered quantitative data about each data structure. The results showed significant difference in performance of each data structure. It was concluded that Vptree is more suitable for searches in smaller populations of neurons and for specific nodes in larger populations, while Kdtree is better for volume searches in larger populations. Octree had highest average search time and memory requirement.
Language: English
Open Access status: green
OpenAlex ID: W2404495620
Cites: 10
Cited by: 2
Related to: 10