Event definitions for Fear Conditioning experiments
Fear conditioning experiments involve conditioned stimuli (CS) and unconditioned stimuli (US). This page extends the BIDS Events specification with the stricter event labels required by the Calibench Fear Conditioning challenge.
In the Calibench Fear Conditioning challenge, event labels distinguish between reinforced CS+ trials, unreinforced CS+ trials, CS- trials, delivered US events, omitted US events, and CS- matched US timepoints.
The events.tsv file MUST use the event labels defined below. These labels are part of the benchmark contract and are used by currently linked methods and importers.
Event labels
Stimulus |
Reinforced? |
Naming |
Description |
|---|---|---|---|
|
yes |
|
A reinforced conditioned trial. |
|
no |
|
An unreinforced conditioned trial. |
|
no |
|
The non-conditioned stimulus. |
|
|
The event in which an unconditioned stimulus, or threat, was present. |
|
|
|
The timepoint where the US would have appeared in an unreinforced CS+ trial. |
|
|
|
The event where the US would have appeared in CS- trials, helping avoid bias in the comparison. |
Required M-BIDS columns
BIDS task events already define onset and duration. In addition to these BIDS columns, Calibench fear-conditioning events.tsv files MUST contain the following M-BIDS columns:
Column |
Description |
Example |
|---|---|---|
|
Fear Conditioning event label. The value must be one of |
|
|
Experimental phase represented by the row. Acquisition files can contain |
|
Do not rename these columns. Additional columns may be included when available, but the importer expects the required columns and event labels listed here.
Optional columns
Calibench datasets may include the following optional columns. They are useful for interpretation and richer metadata, but they are not required by the Calibench fear-conditioning importer.
Column |
Description |
Example |
|---|---|---|
|
Stimulus identifier. If present, use values documented in the sidecar levels. |
|
|
Hierarchical Event Descriptor tags for the event. |
|
Task entities and phases
Calibench fear-conditioning files use the BIDS task entities task-acquisition and task-extinction.
task-acquisition files may include rows whose task_name is habituation before the acquisition trials. task-extinction files use task_name value extinction.
When stimulus_name is present, use dataset-specific values documented in the event sidecar. Common Calibench values include:
Value |
Description |
|---|---|
|
Geometric CS+ stimulus. |
|
Geometric CS- stimulus. |
|
Electrical unconditioned stimulus. |
Example events.tsv
An example task-acquisition events.tsv file with optional stimulus_name and HED columns looks like this:
onset duration event_type stimulus_name HED task_name
10.0 8.0 CSpu square Sensory-event,Visual-presentation,Experimental-condition/CS-plus,Experimental-condition/Unreinforced habituation
17.5 0.0 USo square Experimental-condition/CS-plus,Experimental-condition/Unreinforced,Experimental-condition/Expectation-violation habituation
41.5 8.0 CSm diamond Sensory-event,Visual-presentation,Experimental-condition/CS-minus habituation
49.0 0.0 USm diamond Experimental-condition/CS-minus,Experimental-condition/Expectation-violation,Somatosensory-stimulation habituation
134.4995 8.0 CSpr square Sensory-event,Visual-presentation,Experimental-condition/CS-plus,Experimental-condition/Reinforced acquisition
141.99950000000013 0.4253501000348478 USp shock Sensory-event,Somatosensory-stimulation,Aversive-stimulus acquisition
Interpretation notes
CSpr and CSpu both refer to CS+ trials, but they distinguish whether the trial was reinforced or unreinforced.
USp marks a delivered unconditioned stimulus.
USo marks the omitted US timepoint in an unreinforced CS+ trial.
USm marks the matched timepoint where a US would have appeared during a CS- trial. This allows methods to compare responses at equivalent timepoints and helps avoid bias in the benchmark comparison.