A computer image processing system for quantification of zebrafish behavior.

Area of Science:

• Ethology
• Bioimaging
• Computational Biology

Background:

• Zebrafish (Danio rerio) are widely used models in developmental biology.
• Quantifying complex behaviors like schooling in fish requires advanced analytical tools.

Purpose of the Study:

• To develop and validate a computer image processing system for quantifying zebrafish behavior.
• To enable real-time, high-resolution tracking and analysis of individual and social behaviors in zebrafish.

Main Methods:

• Utilized a CCD camera and graphic I/O board for image acquisition at high spatial (256x256 pixels) and temporal (10 frames/s) resolutions.
• Implemented a skipping search method for high-speed data analysis and developed algorithms to separate occluded images of multiple fish.
• Defined "chasing" behavior based on inter-fish distance, angle, and approach dynamics.

Main Results:

• Successfully tracked and quantified the behavior of both newborn (1.8 cm/s) and adult (7.2 cm/s) zebrafish.
• Observed and quantified schooling behavior, with a 37% chasing ratio in pairs of zebrafish, significantly higher than in pairs of different fish (<20%).
• Demonstrated the system's capability to analyze social interactions, such as one fish chasing another.

Conclusions:

• The developed image processing system provides a powerful tool for quantitative assessment of zebrafish behavior.
• This system facilitates objective scoring of schooling behavior and social interactions in multiple fish.
• Enables deeper insights into the ethology and behavioral ecology of zebrafish models.