Grant: Alliance Grants

Our project addresses gaps in current essential activities involving research animals: (1) rodent models are still the most common for biomedical and pharmacological studies related to disease, including highly infectious agents and toxic substances. Nonetheless, studies involve direct contact with experimental animals with risk mitigation by personal protective equipment. This is expensive and creates continuing health risk; (2) experimental data collection to study rodent health status, behaviour and rodent training for behavioural tasks are always carried out in specialized and expensive apparatus, separate from their homecage, usually in another dedicated room. This approach is a time consuming and labour intensive process and is a major constraint in collecting and managing large volumes of data both in density and in extension within behavioural space. Rodent handling causes inevitable stress which can be a confound in many studies. In addition, it reduces the reproducibility of finding due to unavoidable human error during both animal training and data collection and analysis. We propose to develop a product which is a digital cage, an automated system for animal behaviour data collection and analysis in their homecage for behavioural phenotyping using artificial intelligence. Our system is based on a low-cost combination of video camera and sensors that make our system highly modular and scalable for automated and remote high-throughput data collection and analysis. Our system can be easily adapted to existing animal facility infrastructure to automatize husbandry processes and also to carry out large scale long--term monitoring of brain activity and animal behaviours, and automated data storage and analysis for drug testing or other research. We anticipate that smart homecage platforms for high-throughput longitudinal monitoring of rodents behaviour and Brain Activity will be useful for a wide range of studies in basic and translational biological science, and in particular, it will be useful for studying behavioural phenotypes in genetic models of human disorders, and for relating behavioural abnormalities to their underlying genetic and neural causes.