Respiration (resp)
Following the BIDS Physiological recordings specification, M-BIDS uses the following additional Calibench requirements for respiratory data.
In addition to the BIDS _physio.tsv.gz / _physio.json pair, respiration data MUST use the recording label resp and should be organized as follows:
physio/
├── sub-<participant_label>_task-acquisition_recording-resp_physio.json
└── sub-<participant_label>_task-acquisition_recording-resp_physio.tsv.gz
Column names in the resp data files MUST be timestamp followed by resp for experiments involving single recordings. For different types of respiratory measurements, use timestamp followed by resp1, resp2, resp3, and so on.
These fields MUST be included in the JSON sidecar file:
<matches>_recording-resp_physio.json
Additional Metadata Fields
Key |
Description |
Data Type |
Possible Values |
|---|---|---|---|
|
Type of sensors used for the respiratory measurements |
Text |
Bellows, Cushion, Strain Gauges,… |
|
Placement of the respiratory measurement device |
Text |
Chest |
Example TSV Data File
Compressed *.tsv.gz files are headerless. The columns are defined by the JSON sidecar.
0.0 0.0347900390625
0.0004999999999881766 0.03509521484375
0.0009999999999763531 0.0347900390625
. . .
In this example, the sidecar Columns value is ["timestamp", "resp"].
Example JSON Sidecar
{
"Columns": [
"timestamp",
"resp"
],
"Manufacturer": "Biopac Systems",
"ManufacturersModelName": "ECG100C",
"DeviceSerialNumber": "1711008598",
"SoftwareVersion": "Biopac AcqKnowledge 5.0.2",
"StartTime": 0.0,
"PhysioType": "generic",
"timestamp": {
"LongName": "Time",
"Description": "a continuously increasing identifier of the sampling time registered by the device",
"Origin": "System startup",
"Units": "s"
},
"resp": {
"Description": "Respiratory Recording",
"SensorType": "Belt",
"Placement": "Chest",
"Units": "V"
},
"SamplingFrequency": 2000.0,
"TrimPoints": [
882.7075,
1530.7065
],
"Duration": 647.999
}