Clarify raw and computed physiological data definitions#49
Clarify raw and computed physiological data definitions#49
Conversation
I'm not sure if a corresponding change should be made to modalities.yaml, I'm having trouble finding where the latter table is used by the compiler macros.
Incorporate physio BEP section 4 to physiological-recordings.md
Add physio as a datatype
Some checks on formatting (particularly for how I entered the allowable values) might be necessary here, I also added some comments on the BEP document to flag potential issues or changes we could make going forward.
Removed example file structure for physiological recordings.
Add MeasureType descriptions section to documentation
Removed unnecessary code blocks and clarified recommendations for recording physiological data.
Added descriptions for various MeasureTypes related to physiological recordings.
Updates to src/modality-specific-files/physiological-recordings.md
Rebasing after Eyetracking BEP got merged
Some checks on formatting (particularly for how I entered the allowable values) might be necessary here, I also added some comments on the BEP document to flag potential issues or changes we could make going forward.
Fix some rebasing issues
Added clarification on raw vs computed physiological data and recommendations for data recording. Closes #11
| Examples of computed metrics are heart rate, ... | ||
| Some modalities could be both, e.g. blood pressure directly measured vs computed from PPG signals. | ||
|
|
||
| In the following document, we broadly categorise any computed metric as non-raw data, although currently |
There was a problem hiding this comment.
It concerns me that this could be interpreted as non-raw = BIDS derivative, so we should maybe explicitly clarify that this is not the case.
| To complicate the matter, certain devices allow analog signal processing operations. | ||
|
|
||
| Examples of directly measured physiological fluctuations include respiration, cardiac pulse, or eye-tracking. | ||
| Examples of computed metrics are heart rate, ... |
There was a problem hiding this comment.
Should more examples be specified here other than HR ?
There was a problem hiding this comment.
Yes, I think EDA is a nice one! Do we use "proc" to reflect e.g. phasic vs tonic components?
|
@smoia points out that we recommend saving the 'rawest' data possible, without clarifying what we mean by this... We could change our wording to minimally processed? |
|
A lingistic clarification I suggest: BIDS: raw vs. derivative A transformation of the measured signal --> use proc |
|
@ShiyuBelleWang suggests strongly recommending proc for sparse data |
|
Mary: BIDS: raw vs. derivative where processed relates to a computed metric or transformation of the data EDA tonic and phasic are transformation due to filtering and bandwidth BP acquired on-skin is unprocessed, BP coming from Pulse is processed |
4 participants
Added clarification on raw vs computed physiological data and recommendations for data recording.
Closes #11