String-pulling

String-pulling analyzes video recordings of mice performing hand-over-hand string-pulling to quantify kinematic and statistical features of bilateral skilled forelimb movements for motor phenotyping and study of neurological disorders.

Key Features:

• Task Description: The string-pulling task involves hand-over-hand, repetitive rhythmic forearm movements that assess bilateral skilled hand movements and sensorimotor integration.
• Implementation: Implemented in Matlab®.
• Global motion analysis: Performs optical flow estimation and computes descriptive statistics, principal component analysis (PCA), and independent component analysis (ICA) for movement characterization.
• Temporal complexity metrics: Computes temporal measures including Fano factor, entropy, and Higuchi fractal dimension to quantify movement variability and complexity.
• Image segmentation and tracking: Uses image segmentation and object-tracking heuristic algorithms to independently track body, ears, nose, and forehands.
• Kinematic parameter estimation: Estimates body length, body angle, head roll, yaw, pitch, movement paths, and hand speed from tracked objects.
• Comparative analysis support: Supports analysis of strain-specific postural and skilled-hand kinematic differences.

Scientific Applications:

• Motor phenotyping: Quantitative phenotyping of rodent motor behavior based on kinematic and statistical measures.
• Strain comparison: Comparative analysis of C57BL/6 and Swiss Webster mice to identify postural and skilled-hand kinematic differences.
• Neurological disease models: Assessment of motor deficits in models of Parkinson’s, Huntington’s, Alzheimer’s, and other motor-related disorders.
• Therapeutic and genetic studies: Support for therapeutic drug development and investigation of genetic bases of neural systems underlying behavior.
• Natural behavior analysis: Analysis of unconditioned string-pulling behavior to study sensorimotor integration and spontaneous motor performance.

Methodology:

Implemented in Matlab® and applies optical flow estimation, descriptive statistics, PCA, ICA, Fano factor, entropy, and Higuchi fractal dimension, together with image segmentation and object-tracking heuristic algorithms to independently track body, ears, nose, and forehands and estimate kinematic parameters such as body length, body angle, head roll, yaw, pitch, movement paths, and hand speed.