Open JacobCavon opened 3 months ago
June near end of month Peter has to give update.
Goals for end of quarter: Built software, Recruited patient and apply stimulus w/ software, Streamline data preprocessing
Subject data storage: must be coded (the ones exported will have patient identifiers removed), most complicated part is consent because we're doing studies on unconscious patients eventually (not really relevant to us), must be coded and we have the right to hang onto it in coded form, no requirement to provide data to individual/individual's next of kin, can share with other investigators
RedCap data storage option, free and unlimited data storage. Deidentification files won't go into RedCap, these will be stored in a coded key that allows us to actually review patient data.
From Dave: see how likely it is that we can put our data directly into RedCap without having to generate some other third party key (figure out how to automatically do it). If this is not something we can do on the timescale of the next few weeks we need to have another conversation
How likely/easy to automatically put data into RedCap? Figure out
20-30 Hz range is what we're interested in, but recorded at 256 Hz. So can def downsample to make file size smaller. For range of time maybe just isolate an hour
Rewrite our function to randomize the order of administration of the 3 different stimuli types.
UPDATE: rewrote new functions to administer word, sentences, and beep in random order
Klassen paper has phrases to use. Look into Rodhika's stuff for corpus of words (OSF).
Go to Harborview in-person sometime next week to checkout hardware. Peter is still concerned about identifying/labeling stimulus administration start time.
Where will sensitive data be kept? Fortunately not very limited by EEG. Peter is thinking of getting a study computer/laptop. Also considering Redcap? https://www.lib.washington.edu/dataservices/tools/storage-options Figure out how to get data in there