Genentech

Locate Scans SETUP 423 IS THE NUMBER OF Follow Up Scans on Adams List 99 of them were not on the file system, a third of these have not been retrieved 100 weren't processed because of spacecd /data/pelletier1/genentech/scripts 231 Completed by Alyssa 211 Left for Gina

• cd /data/pelletier1/genentech/scripts
• del ipython notebook
• Open Genentech_FINAL.ipynb
• Year 8/9 data from adams list and the skyra baseline scans: /data/pelletier1/genentech/csv_xlsx/EPIC _GNE_Y8-9_MRI_List.xlsx

RETRIEVE FROM BILL This script checks for all the missing scans from yr8/9 and skyra baseline from mspacman - That list was sent to bill to retrieve - Once Bill retrieved it, I ran ms_get_scanner_data with the EPIC study flag - Copy mse# generated from ms_get_scanner_data into /scratch/henry_temp/PBR/dicom

LOCATE MISSING DATA IN PBR

• Find the 211 scans that alyssa didn't process! - Look at the mseID's from adams list and compare it to the mse's in alyssa's list to see which ones have not been completed Adams list: /data/pelletier1/genentech/csv_xlxs/EPIC _GNE_Y8-9_MRI_List.xlsx - Alyssa's list: /data/pelletier1/genentech/csv_xlsx/genentech_cb_and_skyra_data_WITHMSE_FINAL.xlsx - - Find out what is missing and create the new CSV with all mseID's, columns designate missing versus non-missing! - New Master List: /data/pelletier1/genentech/csv_xlsx/yr89-Gina-notdone_FINAL.xlsx

• Make a text file: /data/pelletier1/genentech/missing.txt, that do not have niftis generated in /data/henry7/PBR/subjects//nifti

• In terminal run the command to create niftis and ratios in pbr for the missing mseIDs: pbr-sge -s /data/pelletier1/genentech/missing.txt --pbrargs "-w nifti"

•   pbr-sge -s /data/pelletier1/genentech/missing.txt --pbrargs "-w ratio"
  

• After that was completed, this script checks through the file system to see whether PBR has been run on the scans in mspacman - /scratch/henry_temp/PBR/dicoms/ for whether pbr has been run to create dicoms - /data/henry7/PBR/subject//nifti for whether pbr has been run to create niftis - /data/henry7/PBR/subject//ratio for whether pbr has been run to create ratios

TROUBLE SHOOTING THE MISSING DICOMS, NIFTIS, RATIOS, OH MY!

• Error with original upload - NO DICOMS when running pbr -w dicom
  • ex error message; could not run node: PBR.dicom.a0
  • Check if there are any series in the mspacman website, if there are..try retrieving them by running ms_dcm_qr -t
  • If that doesn't work.. Ask Jason what the problem is
  • If they aren't in mspacman, ask Bill to re-push the exam and run ms_get_scanner_data
• MseID is for a spinal cord scan not a brain scan - NO NIFTI or RATIO when running pbr -w nifti ratio
  • ex error message; Interface MNIAngulation failed to run or could not run node: PBR.ratio.a0
  • Can check using dti_info, (Gina has a script that will do a batch dti_info)
  • Need to retrieve the brain scan with Bill
• Error with ms_pac_man - NO DICOM/NIFTI/RATIO

  • If not all of the DICOMs were brought down that are in the mspacman system

  • Re-run ms_dcm_qr -t

  • If ms_dcm_qr doesn't bring everything down check dti_info on the dicoms folder in /scratch/henry_temp/PBR/dicoms and see if it matches the series descriptions in the mspacman website

  • If you see something missing you can run ms_dcm_qr -t mseID -s

• The DICOMS that were brought down with the mspacman system are not in the heuristic file
  • ex error message; Interface MNIAngulation failed to run or could not run node: PBR.ratio.a0
  • Need a T1 and a T2/FLAIR in order to create a ratio
  • Talk to Anisha about adding a series name to the heuristic file
• Error: command 'antsApplyTransforms' could not be found
  • Add to path: #ANTS export PATH=/netopt/rhel7/versions/ANTs/1.9/antsbin/bin:$PATH export ANTSPATH=/data/henry7/software/ak/ #/data/henry6/keshavan/ANTS/antsbin/ export PATH=/netopt/rhel7/versions/ANTs/1.9/ANTs/Scripts:$PATH
• MT files don't match size
  • For now MT is taken out of the pipeline so we shouldn't get this error
  • However, once MT is back we need to discuss with Bill what the correct MT series pair names are

SETUP RATIOS AND T1s

• See Genentech_FINAL script to move ratio scans into: /data/pelletier1/genentech/GKyr89_ratio/all_ratios_renamed/
• Move T1 aligned scans into: /data/pelletier1/genentech/GKyr89_ratio/T1_aligned_renamed

Histogram Normalization

• Run the following command:
• python /data/pelletier1/genentech/scripts/histogram_normalization_GK.py /data/pelletier1/genentech/GKyr89_ratio/all_ratios_renamed/*
• (python 2 for this script)
• This will generate thresholds, to get the average threshold look at the average 40-50 column threshold
• Copy that threshold and save it to a seperate csv: /data/pelletier1/genentech/average_threshold_genentech_yr89.xlxs

Lesion Registration

• Run the following command:
• cd /data/pelletier1/genentech/GKyr89_ratio/all_ratios_renamed/
• python /data/pelletier1/genentech/GKyr89_ratio/all_ratios_renamed/register_lesion_wrapper_TEST.py /data/pelletier1/genentech/GKyr89_ratio/all_ratios_renamed/*
• (you may have to move the files that are called warper into a new folder to run this - but remember to move them back when it's done)
• (you may have to move the skyra BL scans because they do not have any previous lesions masks - but remember to move them back)
• This script will look at old lesion masks and register the new ones to the old and generate a series of files
• For the skyra BL scans the lesions were created this time around and therefore did not need to be registered

SIENAX for Follow Up

• cd /data/pelletier1/genentech/yr8_sienax

• For non baseline scans you run: python /data/pelletier1/genentech/yr8_sienax

• /run_sienax_y8_GK.py /data/pelletier1/genentech/GKyr89_ratio/all_ratios_renamed/*

SIENAX For Skyra BL

• Antje and Simone edited the lesion masks in Jim
• Esha converted the lesion masks into ROIs
  • mkdir /data/pelletier1/genentech/yr8_sienax/ms****-mse****
  • Grab the aligned T1 image in /data/henry7/PBR/subjects//aligned and put it in /data/pelletier1/genentech/yr8_sienax/ms****-mse****
  • Put the lesion mask in (.nii.gz) in /data/pelletier1/genentech/yr8_sienax/ms****-mse****
  • run the following sienax command in /data/pelletier1/genentech/yr8_sienax/ms****-mse****
  • sienax -lm -r -d -o /data/pelletier1/genentech/yr8_sienax/ms****-mse****

Collect DATA from SIENAX

• cd /data/pelletier1/genentech/yr8_sienax
• (make sure you are using python 2)
• python collect_sienax_data_GKedits.py *
• Data will be saved in a csv called: genentech_yr89_data.csv