• Specifications
• DataManagement
• Discussion

Discussion surrounding the folder structure of in vitro data.

Current State of Organization

Raw Data Collection:

• CMULTIS001-1
  • Ultrasound: DICOM images using trial number and Subj ID
  • Data: TDMS files using trial number and Subj ID
  • Configuration: .cfg sensor and state files using trial number and Subj ID (also the full subject xml and config file)
  • MRI: NIFTI files of all segments using Subj ID and segment identifier (WA, UL, LL), includes T1-weighted (T1) and T1-weighted with fat supression (FS) [6 NIFTIs)
  • CT: NIFTI files of both extremities using Subj ID and limb identifier (WA, WL) [2 NIFTIs]
  • Video: TBD
  • Mach1 (Tissue Testing): TBD
  • Vic3D (Tissue Testing): TBD

Data Analysis folders appended:

• File Association Folders

  • DataOverview: Same as in vivo - Trial number

  • FileAssociation: Same as in vivo - Trial number

  • TimeSynchronization: Same as in vivo - Trial number

• Registration Folders
  • Registration
    • Segmentation Instance (i.e. 01, 02, etc., include XML descriptor in segmentation instance root folder - TBD)
      • CT
        • Segmentation: Contains fiducial marker STLs and label map used to create STLs. Named using CT name with fiducial label (CMULTIS007-1_F1_CT.stl). Labels named using CT name with -label appended to the end (CMULTIS007-1_WL_CT-label.nii).
        • USPositions: XML for each segment with the ultrasound positions in the CT coordinate system (CMULTIS007-1_LA_CT_US_TIMESTAMP.xml)
      • MRI
        • Segmentation: Same as CT except with MR (CMULTIS007-1_F1_MR.stl)
        • USPositions: Same as CT except with MR (CMULTIS007-1_LA_CT_TIMESTAMP.xml)

      • QualityCheck: Contains registration summary and errors as .txt files named with the SubjID and segment identifier as well as the comparison it makes (digitized spheres - DG, CT, or MR) and 'MQC' for marker quality check. (e.g. CMULTIS007-1_LA_DG_CT_MQC.txt)

• Tissue Thickness Folders
  • Tissue Thickness

    • UltrasoundManual: Same as in vivo - named with trial number

      • ThicknessPNG
    • CTManual_zProbe: Named with trial number. Include Registration ID in the xml?
      • ThicknessPNG
    • MRManual_zProbe: Named with trial number. Include Registration ID in the xml?
      • ThicknessPNG

Questions/Thoughts

Registration is a splitting point for the thickness data (CT and MRI) and the model development. Since we may have multiple registrations completed by different people or at different times, I would suggest introducing a registration ID instead of just a sub-directory that is currently in place. This registration ID would be placed in the filename or metadata which is outlined below.

-Registration

• -CT
  • -Segmentation

    • R01_CMULTIS007-1_WA_CT_H1.stl
      R01_CMULTIS007-1_WA_CT-label.nii

    -USPositions

    • R01_CMULTIS007-1_LA_CT_TIMESTAMP.xml

    -zProbeSlices

    • 004_CMULTIS007-1_UL_AP_A-2_CT_R01.vtk

  -MRI: same changes as above in CT

  -QualityCheck

  • R01_CMULTIS007-1_LA_DG_CT_MQC.txt

-TissueThickness

• -CTManual_zProbe
  • Each XML will have an additional field that includes the registration ID
  -MRManual_zProbe
  • Same as CT tissue thickness

Query workflow
Ultrasound indentation data is accessed using the trial number.

• - This includes raw data and derivative data with trial number (excludes CT and MRI as well as fiducial segmentations and position xmls from registration).

Registration data (data excluded from ultrasound indentation should be accessed using the registration ID. (Working on an XML to test for query system)