From mazes to automation: Modernizing working memory research in animal models

Area of Science:

• Cognitive Neuroscience
• Animal Behavior
• Neuroscience Technology

Background:

• Working memory (WM) is crucial for adaptive behavior.
• Traditional rodent models (e.g., Radial Arm Maze, T-Maze) for WM research are manual, limiting precision and scalability.
• Existing methods present challenges in high-throughput data collection and experimental rigor.

Purpose of the Study:

• To explore how automated technologies (touchscreens, VR, AI) can advance rodent WM research.
• To introduce a framework for evaluating WM tasks based on Scalability, Precision, and Neural Compatibility.
• To highlight modern WM tasks enabled by automation.

Main Methods:

• Review and refinement of automated technologies in WM research.
• Development of a novel evaluation framework for WM tasks.
• Analysis of modern automated WM tasks like the Pulse-Based Accumulation Task and Trial-Unique Nonmatching-to-Location (TUNL).

Main Results:

• Automated technologies enable high-throughput WM testing with enhanced precision in rodent models.
• The proposed framework effectively evaluates WM tasks on key metrics.
• Modern automated tasks offer more precise WM assessments compared to traditional methods.

Conclusions:

• Technological advancements significantly improve the quality and efficiency of WM research.
• Automation facilitates the development of sophisticated WM paradigms.
• These progressions contribute to a more unified understanding of the neural basis of working memory.