Unraveling cognitive traits using the Morris water maze unbiased strategy classification (MUST-C) algorithm

Brain Behav Immun. 2016 Feb:52:132-144. doi: 10.1016/j.bbi.2015.10.013. Epub 2015 Oct 24.

Abstract

The assessment of spatial cognitive learning in rodents is a central approach in neuroscience, as it enables one to assess and quantify the effects of treatments and genetic manipulations from a broad perspective. Although the Morris water maze (MWM) is a well-validated paradigm for testing spatial learning abilities, manual categorization of performance in the MWM into behavioral strategies is subject to individual interpretation, and thus to biases. Here we offer a support vector machine (SVM) - based, automated, MWM unbiased strategy classification (MUST-C) algorithm, as well as a cognitive score scale. This model was examined and validated by analyzing data obtained from five MWM experiments with changing platform sizes, revealing a limitation in the spatial capacity of the hippocampus. We have further employed this algorithm to extract novel mechanistic insights on the impact of members of the Toll-like receptor pathway on cognitive spatial learning and memory. The MUST-C algorithm can greatly benefit MWM users as it provides a standardized method of strategy classification as well as a cognitive scoring scale, which cannot be derived from typical analysis of MWM data.

Keywords: Cognitive score; Hippocampus; Learning and memory; Machine learning; Morris water maze; SVM; Spatial learning; Spatial resolution; Strategy.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Algorithms*
  • Animals
  • Cognition / physiology*
  • Hippocampus / physiology
  • Male
  • Maze Learning / physiology*
  • Memory / drug effects*
  • Mice
  • Mice, Inbred C57BL
  • Mice, Transgenic
  • Space Perception / physiology
  • Spatial Learning / physiology*
  • Support Vector Machine
  • Swimming / physiology