Occurrence density index for behavior classification of zebrafish larvae

Abstract

Larval zebrafish are proving to be promising subjects for research that is not subject to legal requirements. However, the behavior of this animal has not been fully explored by researchers yet. This paper proposes a new occurrence-density index (ODI) for behavioral analysis and classification of zebrafish larvae. The ODI is identified through a multistage process that includes (i) mapping of the testbed arena into a virtual arena, (ii) projection of the original object's trajectory into the virtual arena, (iii) assessment of the occurrence-density, and (iv) calculation the proposed index. The results obtained from this pilot study clearly demonstrated a promising ability to classify behaviors of zebrafish larvae. The ODI will therefore not only provide a new classification feature to the known set of features for fish behavior analysis but also explain and distinguish different behavioral traits.