Rearing event log
Every rearing event with onset time, offset time, duration, and classification (wall-supported vs. unsupported).
CSVJSON
← Capabilities
Rearing Detection
Automated vertical activity detection from standard video
Quantify rearing frequency, duration, and type — wall-supported versus unsupported — without manual scoring or specialized side-view cameras.
94%
Rearing event detection accuracy
2
Rearing subtypes classified
33ms
Temporal resolution per frame
0
Manual scoring required
The problem
Rearing is a high-value measure that is expensive to score
Rearing frequency is a primary indicator of exploration, anxiety, and dopaminergic function. Yet most labs still rely on manual scoring — a trained observer watches video and counts events, introducing fatigue bias and inter-rater variability.
- Manual rearing counts take 2-3x real time and introduce observer fatigue bias
- Inter-rater reliability for rearing classification rarely exceeds 85%
- Wall-supported versus unsupported rearing — a key distinction for anxiety — is often collapsed into a single count
The solution
Pose-based rearing classification at frame level
ConductVision uses body posture features from pose estimation to detect rearing events automatically. The system distinguishes wall-supported rears (forepaws on wall) from unsupported rears (free-standing), and reports onset, offset, and duration for each event.
- Automated detection from overhead video — no side camera or manual scoring needed
- Wall-supported and unsupported rearing classified separately for anxiety phenotyping
- Frame-level onset/offset timestamps enable precise bout duration analysis
Endpoints
Rearing behavior dependent variables
Rearing frequency summary
Total rearing count, mean bout duration, wall-supported ratio, and time-binned rearing frequency across the session.
CSV
Annotated rearing video
Original video with rearing events highlighted — bounding box color indicates rearing subtype.
MP4
Applications
Paradigms using rearing as a dependent variable
Open field vertical activity
Rearing frequency in the open field is a standard measure of exploratory drive. Unsupported rears indicate higher exploration motivation than wall-supported rears.
Measures
- Total rearing count
- Unsupported rear ratio
- Rearing frequency per minute
Elevated plus maze rearing
Rearing in open arms versus closed arms of the EPM provides an additional anxiety index beyond arm entry counts.
Measures
- Open arm rearing count
- Closed arm rearing count
- Rearing ratio (open/total)
Dopaminergic drug screening
Rearing is dose-dependently affected by dopamine agonists and antagonists. Automated counting enables high-throughput dose-response curves.
Measures
- Rearing count by dose
- Bout duration by dose
- Latency to first rear
Within-session habituation
Declining rearing frequency across time bins within a session indexes habituation to novelty — a learning measure independent of locomotion.
Measures
- Rearing per 5-min bin
- Habituation slope
- First vs. last bin ratio
Compared to typical systems
How ConductVision differs
| Feature | ConductVision | Typical systems |
|---|---|---|
| Automated rearing detection | Yes — pose-based, no side camera | Manual scoring or side-view camera required |
| Wall-supported vs. unsupported | Classified automatically | Rarely distinguished |
| Temporal resolution | Frame-level (33 ms) | Event-level (observer reaction time ~500 ms) |
| Inter-rater variability | Zero — deterministic algorithm | 10-15% typical inter-rater disagreement |
| Throughput | Real-time, unlimited sessions | 2-3x real time per session per scorer |
Related capabilities
Behavior Classification
ML-powered frame-level behavioral classification with temporal smoothing and custom classifier training.
Pose Estimation
Multi-keypoint body part tracking — head, paws, tail, body center — with pretrained and custom model support.
Anxiety and Depression Index Scores
Composite behavioral indices with Z-score normalization across anxiety and depression test batteries.
Stop counting rears by hand
Run a pilot recording through ConductVision and compare automated rearing counts against your manual scores.
