A generalised framework for detailed classification of swimming paths inside the Morris Water Maze

Abstract

The Morris Water Maze is commonly used in behavioural neuroscience for the study of spatial learning with rodents. Over the years, various methods of analysing rodent data collected during this task have been proposed. These methods span from classical performance measurements to more sophisticated categorisation techniques which classify the animal swimming path into behavioural classes known as exploration strategies. Classification techniques provide additional insight into the different types of animal behaviours but still only a limited number of studies utilise them. This is primarily because they depend highly on machine learning knowledge. We have previously demonstrated that the animals implement various strategies and that classifying entire trajectories can lead to the loss of important information. In this work, we have developed a generalised and robust classification methodology to boost classification performance and nullify the need for manual tuning. We have also made available an open-source software based on this methodology.

Constitutive differences in glucocorticoid responsiveness are related to divergent spatial information processing abilities

“…In brief this approach divided the trajectories into segments, which were classified into different classes of behavior. Results from this analysis led to a detailed categorization of swimming paths and allowed the detection of mixed strategies within a single trial (Gehring et al, 2015; Vouros et al, 2018). Eight different behavioral strategies were considered: Thigmotaxis, when an animal was swimming close to the walls of the arena; Incursion, when an animal started to move towards the inward locations of the arena; Scanning, when an animal was swimming randomly in the whole arena (Graziano et al, 2003); Focused Search, when an animal focused its search on a region of the arena different to the platform; Chaining Response, when an animal was swimming at the distance of the platform from the arena wall (Wolfer and Lipp, 2000); Sell Orienting, when an animal performed a loop while swimming, thereby orienting itself inside the arena (Graziano et al, 2003); Scanning Surroundings strategy, when an animal crossed the platform or the region around it; Scanning Target strategy when an animal performed a Focused Search within the platform region.…”

“…On the other hand. Self Orienting and Scanning Target strategies are considered as high-level strategies during which animals use the visual cues around the pool to orient themselves and find the platform (Gehring et al, 2015; Janus, 2004; Vouros et al, 2018).…”

“…Moreover, as the CORT elevation displayed by the High line in response to a stress challenge (Walker et al, 2017) does not reach the high CORT levels found to induce cognitive impairments in a Morris water maze task (Salehi et al, 2010), we hypothesized that the High line would benefit from the endogenous higher glucocorticoids response and display superior spatial information processing capacities (particularly long-term memory) as compared to the Low and Inter CORT lines. To gain insight into the effects of stress on spatial learning abilities, we expanded the analysis of the traditional parameters from the Morris water maze by implementing a novel classification method that captures detailed aspects of swimming performances and exploratory strategies (Gehring et al, 2015; Vouros et al, 2018).…”

Sodium nitroprusside prevents the detrimental effects of glucose on the neurovascular unit and behaviour in zebrafish

“…Using previous rodent based methods developed for Morris water maze (Wolfer & Lipp, 2000;Graziano et al, 2003;Gehring et al, 2015;Illouz et al, 2016;Vouros et al, 2018), we quantified four different features from the swimming trajectories of zebrafish. Coordinates of the swimming trajectories were extracted from the Viewpoint-…”

Pathfinder: open source software for analyzing spatial navigation search strategies

“…One likely explanation is that it is not a standard feature of 23 commercially-available software packages, therefore requiring time and programming 24 experience to execute. Groups that have performed strategy analyses have developed their 25 own software, using either a predefined parameter-based approach, like ours, or machine 26 learning algorithms that classify based on user input 8,11,12,15-20, 25,41 . Since most previous 27 approaches have not been developed into freely-available software packages, Pathfinder may 28 enable more widespread adoption of strategy analyses.…”