ADRD_analysis

Scripts and functions to preprocess and analyze ADRD data.

files needed:

• probe_corrected.SD (the correct probe to be placed in snirf files)
• roi_coords_dem.csv (file that maps channel values to different brain regions)

scripts needed:

• Replace_Probe.m
• Rename_snirfs_and_make_BIDS_folders.m
• rename_stims.m
• generate_DQRs.py (python/ cedalion)
• analysis script (or can run analysis in Homer GUI) *need Cedalion environment created

Preprocessing Scheme (updated for auto) 1 run Replace probe script ( Replace_Probe.m ) 2. run rename snirfs and create bids folders script 3. run rename stims script 4. generate DQR's (see below) 5. Run analysis in Homer GUI or script based

Steps to generate DQRs:

1. open a scc login node session
2. Run the following commands: 2.2) module load miniconda 2.3) conda activate cedalion 2.4) cd /projectnb/nphfnirs/s/datasets/U01_ADRD/code/ADRD_analysis/

2.5) python3 -c "from generate_DQRs import run_DQR_all_tasks; run_DQR_all_tasks( subj = '06', task_lst = ['Cloudy', 'GEN1', 'GEN2', 'PERS1', 'PERS2', 'Presto'], snr_thresh=10, sd_thresh= [0, 45], amp_thresh=[0, 1e7])"

NOTE: change the subject number, task_lst if it varies, snr_thresh, sd_thresh, and amp_thresh to values you want to use.

Function description: ''' Function that loops through each task in task_lst for a single subject and generates and saves the DQR plot for each task.

Inputs: 
    subj (string) : subject number (ex '01')
    task_lst (list) : of strings indicating each task name for a subject.
    snr_thresh (int) : threshold for signal to noise ratio
    sd_thresh (list) : list of int class of source detector separation thresholds. 
        first element lower thresh, second element upper threshold. ex [0, 45]. 
    amp_thresh (list) : list of int class of amplitude thresholds. 
    fmin (int) : lower threshold  value for bandpass filter
    fmax (int) : upper threshold value for bandpass filter

'''