ABSTRACT
OBJECTIVE: The freezing response in rats has been used widely as an index of fear in threatening or stressful situations. Considering its importance in basic studies of affective disorders and in drug testing, it is necessary to develop a reliable, automated system for measuring the freezing response accurately. METHODS: We developed the Vision-Based Analysis for Freezing (V-BAFF) system to analyze the freezing response in rats, based on several image-processing modules and motion classification schemes generated from our own experience. We compared the performance of V-BAFF with that of experienced human observers. RESULTS: The level of freezing measured with V-BAFF was highly correlated with that measured by human observers. CONCLUSION: We demonstrated that V-BAFF could yield a bias-free, accurate, and efficient measurement of freezing in rats.