Analyzing Freezing Behavior in Fear Conditioning Experiments Using EthoVision Activity Analysis
Last updated: Nov 01, 2026
Problem
Users want to automatically detect and quantify freezing behavior during fear conditioning experiments using EthoVision XT, but are unsure how to configure the Activity State analysis parameters to accurately identify freezing bouts.
Details
EthoVision XT can detect freezing using its Activity State analysis, which classifies frames as 'mobile' or 'immobile' based on pixel change thresholds between successive frames. Default threshold values may not accurately reflect freezing for a given experimental setup, animal species, or camera configuration. Users may get inconsistent results if the Activity State parameters (mobility threshold, minimum duration) are not tuned to their specific conditions. Relevant for fear conditioning, open field, and other paradigms requiring immobility detection. Affects EthoVision XT across multiple versions.
Solution
1. Open your experiment in EthoVision XT and navigate to the Analysis Profile for your trials.
2. Ensure you are using the 'Activity State' parameter (found under Movement Analysis), which classifies the animal as mobile or immobile based on activity levels.
3. To calibrate the threshold: manually score a representative subset of trials using the Manual Scoring module, marking clear freezing bouts. This creates a reference against which you can compare the automated detection.
4. Adjust the 'Activity' threshold value in the Analysis Profile — a lower threshold means less pixel change is needed to classify the animal as mobile (may over-detect movement); a higher threshold will require more movement (may miss subtle motion).
5. Adjust the minimum duration parameter to set the shortest freezing bout that will be counted (e.g., 1–2 seconds), filtering out very brief pauses.
6. Run the analysis on your manually scored trials and compare the automated output to your manual scores.
7. Iterate on the threshold and minimum duration settings until automated and manual scoring align acceptably.
8. Apply the finalized Analysis Profile to all trials for consistent results.
9. Export the resulting statistics (e.g., total time immobile, number of freezing bouts) to Excel for further analysis.