研究目的
To provide an open-source high-speed infrared videography database for studying the principles of active sensing in freely navigating rodents, focusing on whisker-based tactile navigation and sensorimotor computation.
研究成果
The database provides a valuable resource for studying sensorimotor computation, developmental changes, and cross-species comparisons in active sensing. It supports machine learning for automated analysis and offers insights into adaptive motor control and the role of serotonergic signaling.
研究不足
The dataset does not include data from female animals. Freely behaving experiments have high dimensionality and sampling limitations, but the current dataset with 6,642 videos across multiple conditions is sufficient for reproducible statistics. Variations in animal behavior (e.g., approach angle, whisking kinematics) across trials may affect consistency.
1:Experimental Design and Method Selection:
The study involved designing a behavioral paradigm (gap-crossing task) under infrared light to observe rodents locating a tactile target. High-speed videography was used to capture whisker and body movements. Analytical methods included manual and automated tracking of whiskers and nose positions.
2:Sample Selection and Data Sources:
38 male rats and 10 male mice were used, including wild-type, serotonin transporter knockout rats, and pharmacologically treated rats. Animals were juveniles (postnatal weeks 3-4) or adults, with some subjected to sensory deprivation (single or multi-whisker conditions).
3:List of Experimental Equipment and Materials:
High-speed cameras (PointGrey Flea3 and AVT Pike), infrared backlight, linear actuators, motion sensors, servo motors, polyvinyl chloride panels, custom software in MATLAB, and pharmacological agents like fluoxetine hydrochloride.
4:Experimental Procedures and Operational Workflow:
Animals were habituated and trained in the gap-crossing task. Videos were recorded during exploration sessions, with data acquisition triggered by motion sensors. Human observers manually tracked whisker positions in a subset of videos for ground-truth data.
5:Data Analysis Methods:
Data were stored in .mat and .mp4 formats, analyzed using MATLAB functions and open-source software for image processing and machine learning approaches.
独家科研数据包,助您复现前沿成果,加速创新突破
获取完整内容