PSYCO Pybased SYstem for Chronic monitoring of r

PSYCO: Py-based SYstem for Chronic monitoring of r. Odent Species, an open-sourced python based system for home-cage mouse tracking and image processing Tony L. Fong, Timothy H. Murphy Department of Psychiatry, University of British Columbia, Vancouver, Canada Evaluation Results Activity Level Analysis Over Three Days Duration (s) • Markerless chronic tracking of mice • Automatic chronic tracking and monitoring of mice remain difficult (Chaumont et al. , 2018) (Peleh et al. , 2019) • Behavior recording on a customizable/scalable Pibased setup • Open-sourced offline Python-based analysis tools to chronically track multiple mice Speed (pixels/s) Introduction Ethernet Connection For Remote Access Pi Camera Recording Resolution Rodent Arena Hard Drive Raspberry Pi 3 B+/ Pi 4 Additional Features RFID Implanted Mice USB Hub B Recording Resolution Distance (pixels) A Count Behavior Capture Setup Integration with Deeplabcut RFID Readers IR-LED Fisheye Lens Pi Camera Animals in Video Track Different Rodent Strains in Different Environments Travel Trajectory Analysis of Mice • The total cost to setup one system with 5 RFID reader is less than 700 USD Tracking Pipeline SORT Tracking Yolov 4 Mouse Detections Conclusion • You Only Look Once version 4 (Yolov 4) mice detections • Yolov 4 reaches • Simple Online Real Time (SORT) Tracking • RFID-Sort ID Association RFID-ID Association • PSYCO is a cost-effective and customizable/scalable system to track rodent spices • Mice activity levels are high during dark phase compared to light phase References Chaumont et. al. , (2018). Nature Biomedical Engineering, 93(4), 929– 939. e 6. Peleh et. al. (2019. Journal of Neuroscience Methods 325 (2019): 108323. Acknowledgements: CIHR Canada fund and UBC IMH Marshall Scholars and Fellows Program THE UNIVERSITY OF BRITISH COLUMBIA Acknowledgements: CIHR Canada fund and Canadian Neurophotonics Platform
- Slides: 1