Event Definitions
Task events are saved as events.tsv and events.json files (see the BIDS Events specification).
In addition to the event definitions provided in the BIDS specification, M-BIDS requires event labels to be stored in an event_type column. The exact allowed event labels depend on the experiment type or benchmark challenge. For Calibench fear-conditioning datasets, these labels and the accompanying event columns are part of the importer contract and must be written exactly as documented.
Expected file location
Fear-conditioning event files should be stored together with the corresponding task recording files. The target folder depends on the available physiological modalities.
Available data |
Target folder |
Reason |
|---|---|---|
Eye-tracking only |
|
Eye-tracking-only datasets can store task events and eye recordings in the behavioral folder. |
SCR only |
|
Skin conductance recordings are physiological recordings and should be stored in the physiology folder. |
Eye-tracking and SCR |
|
When physiological recordings are present together, the task recordings and corresponding events are stored in the physiology folder. |
For Calibench BEP020/BEP045 datasets, physiology and eye-tracking recordings are stored together in physio/. The structure for one subject and task should look like this:
sub-<participant_label>/
└── physio/
├── sub-<participant_label>_task-acquisition_events.json
├── sub-<participant_label>_task-acquisition_events.tsv
├── sub-<participant_label>_task-acquisition_recording-scr_physio.json
├── sub-<participant_label>_task-acquisition_recording-scr_physio.tsv.gz
├── sub-<participant_label>_task-acquisition_recording-ecg_physio.json
├── sub-<participant_label>_task-acquisition_recording-ecg_physio.tsv.gz
├── sub-<participant_label>_task-acquisition_recording-resp_physio.json
├── sub-<participant_label>_task-acquisition_recording-resp_physio.tsv.gz
├── sub-<participant_label>_task-acquisition_recording-eye2_physio.json
├── sub-<participant_label>_task-acquisition_recording-eye2_physio.tsv.gz
├── sub-<participant_label>_task-acquisition_recording-eye2_physioevents.json
└── sub-<participant_label>_task-acquisition_recording-eye2_physioevents.tsv.gz
The same pattern is used for task-extinction. A subject may contain only the modalities that were actually recorded, but every included recording must have a matching JSON sidecar.
For an SCR-only fear-conditioning dataset, the structure may look like this:
sub-<sub_id>/
└── physio/
├── sub-<sub_id>_task-acquisition_events.json
├── sub-<sub_id>_task-acquisition_events.tsv
├── sub-<sub_id>_task-acquisition_recording-scr_physio.json
└── sub-<sub_id>_task-acquisition_recording-scr_physio.tsv.gz
If both eye-tracking and physiological recordings are present, the files should be placed under physio/. This placement follows the modality handling described by BEP020 and BEP045 and is supported by the BIDS importer.
Required M-BIDS columns
BIDS task events already define onset and duration columns. In addition to these BIDS columns, Calibench fear-conditioning events.tsv files MUST contain the following M-BIDS columns:
Column |
Description |
Data Type |
Example |
|---|---|---|---|
|
Type of event, expressed as a character string. The allowed values depend on the experiment type. |
String |
|
|
Experimental phase represented by the row. |
String |
|
Optional columns
Additional columns may be included when available or required by a specific dataset, experiment, or benchmark challenge. They must not replace or rename the BIDS columns or required M-BIDS columns listed above.
Column |
Description |
Data Type |
Example |
|---|---|---|---|
|
Name or identifier of the presented stimulus. |
String |
|
|
Hierarchical Event Descriptor tags describing the event. |
String |
|
|
Participant response, when responses are part of the task. |
String |
|
|
Time between stimulus onset and participant response. |
Number |
|
Specific event types
The following generic event types may be used to describe common experimental events.
Event Type |
Description |
|---|---|
|
Start of an experimental block. |
|
End of an experimental block. |
|
Generic participant response to stimuli. |
Specific guidelines are different for different experiment types and are listed below: