Automated facial feature evaluation system to prevent stress of head fixed mice

Abstract

Head fixation of rodents is a widely utilized and important technique that enables laboratories to measure brain activity during behavior, but head fixation can increase stress which affects both behavior and underlying brain activity, as well as animal welfare. It is therefore critical to keep stress levels low, yet it is particularly challenging to assess stress in immobilized, head-fixed rodents. Conventional stress evaluation methods, such as blood corticosterone analysis, are labor-intensive and conducted post hoc, and In situ approaches require experienced personnel and constant vigilance during experiments. In this study, we present MouseCare, an automated software solution for immediate stress detection by real-time facial feature video analysis in head-fixed mice. MouseCare performs on par with or better than conventional stress measures, and is significantly less labor-intensive. This approach enables objective measures of stress that are needed to determine when an experiment can commence, but also when it should be stopped. We conclude that MouseCare offers a cost-effective and easily implementable strategy to manage stress levels in rodents that can increase data quality and improve animal welfare.