A simple method that doubles learning speed for mice in touchscreen-based visual discrimination.

Touchscreen-based operant learning systems are widely used in behavioral neuroscience to assess cognitive function across various animal models, including mice. However, conventional training protocols for visual discrimination (VD) tasks in mice often require extended training periods. We developed modified touchscreen training protocols that enhance learning efficiency in VD tasks. Our method requires mice to perform two or three consecutive screen touches to confirm their stimulus selection. If inconsistent responses occur during these touches, the trial is reset. This approach significantly improved task acquisition speed. Additionally, we incorporated visual feedback after incorrect choices, which further enhanced learning speed and task accuracy. Using this protocol, mice rapidly acquired both initial VD learning and reversal learning, even with multiple randomly presented stimulus pairs. Furthermore, in serial reversal tasks, mice exhibited improved performance across sessions, indicating rule acquisition and cognitive flexibility. These findings demonstrate that simple protocol adjustments can significantly improve touchscreen-based cognitive assessments in mice, enabling more cognitively demanding behavioral experiments in preclinical research.