Stride metrics per paw
Stride length, stride frequency, stance duration, swing duration, and duty cycle computed independently for each paw across the session.
CSV
← Capabilities
Gait Analysis
Automated stride detection and paw-level locomotor scoring
Extract stride length, stance/swing duration, and paw-specific metrics from standard video — no specialized treadmill required.
4paws
Independently tracked
HMM
Stride detection method
Standard
Camera — no specialized equipment
Automated
Scoring pipeline
The problem
Gait scoring requires expensive equipment or tedious manual analysis
Dedicated gait analysis systems (CatWalk, DigiGait) cost tens of thousands of dollars and require animals to walk on specialized treadmills or glass platforms. Manual gait scoring from video is time-intensive, subjective, and typically limited to gross measures. Neither approach scales to large cohorts or longitudinal studies.
- CatWalk and DigiGait systems require capital investment of $50,000+ and specialized apparatus
- Manual frame-by-frame gait scoring takes 30-60 minutes per session and has high inter-rater variability
- Specialized equipment constrains the animal to unnatural locomotion contexts (treadmill, glass walkway)
The solution
Stride metrics from standard video using hidden Markov models
ConductVision uses HMM-based stride phase detection applied to paw keypoint trajectories from standard video. Stance and swing phases are identified for each paw independently, yielding stride length, cadence, duty cycle, and interlimb coordination — all from a standard overhead or side camera.
- HMM-based stride phase detection identifies stance and swing from paw keypoint trajectories
- Four paws tracked independently — stride length, cadence, duty cycle, and base of support per limb
- Balance beam traversal scoring included — foot slips, traversal time, hindpaw placement accuracy
Endpoints
Locomotor dependent variables at paw-level resolution
Phase coordination and base of support
Interlimb phase relationships, regularity index, and dynamic base of support. Detects coordination breakdown in neurological models.
CSV
Balance beam scoring
Traversal time, foot slip count, hindpaw placement relative to beam center, and hesitation events for beam walk protocols.
CSV
Applications
Motor deficit paradigms quantified from standard video
Stroke recovery motor assessment
Track progressive recovery of gait symmetry after unilateral ischemia. Stride length asymmetry and interlimb coordination quantified longitudinally from standard cage or open field video.
Measures
- Stride length asymmetry
- Interlimb coordination index
- Swing duration ratio
Parkinson's model locomotor deficit
Detect shuffling gait, reduced stride length, and increased stance time in 6-OHDA or MPTP models. Quantify L-DOPA response and wearing-off dynamics.
Measures
- Stride length
- Cadence
- Stance/swing ratio
Spinal cord injury progression
Longitudinal tracking of hindlimb locomotor recovery after spinal cord injury. Base of support, regularity index, and stepping pattern quantified across weeks.
Measures
- Hindlimb base of support
- Regularity index
- Plantar stepping frequency
Drug-induced motor effects
Detect and quantify ataxia, catalepsy, and motor slowing from acute drug administration. Per-paw metrics reveal lateralized effects missed by gross locomotor measures.
Measures
- Gait regularity
- Paw drag duration
- Locomotor speed variability
Compared to typical systems
How ConductVision differs
| Feature | ConductVision | Typical systems |
|---|---|---|
| Equipment required | Standard camera | Specialized treadmill or glass walkway ($50K+) |
| Per-paw resolution | 4 paws independent, HMM stride detection | Varies by system |
| Locomotion context | Free walking — open field, beam, corridor | Constrained to apparatus |
| Paradigm flexibility | Open field, beam walk, corridor, any arena | Single apparatus only |
| Cost | Included in ConductVision license | $50,000+ dedicated system |
Related capabilities
Pose Estimation
Multi-keypoint body part tracking — head, paws, tail, body center — with pretrained and custom model support.
Precision Tracking
High-resolution 30 fps tracking that captures sub-second behavioral events conventional systems miss.
Behavior Classification
ML-powered frame-level behavioral classification with temporal smoothing and custom classifier training.
Analyze gait from your existing recordings
Upload video from any camera angle and extract per-paw stride metrics without specialized hardware.
