/
Approach for MRI

Approach for MRI

Owned by Paul Macey (admin)
Last updated: May 06, 2022 by Paul Macey


These are the steps to set up and run an analysis in CSPM, which is installed as a toolbox in SPM. These steps assume that the CSPM software is already set up. CSPM was originally created for MRI data, but now physiology and discrete measures (surveys, vitals) can be analyzed.

Subjects

Add a new subject to CSPM.

MRI Data

  1. Store data in structured folder structure ("Study Data").
  2. Define a Protocol id for each set of scans.
    • Note that "examno" and ProtocolID" and "PatientID" are all sometimes used in place of session id.
  3. Convert MRI DICOM/PARREC files, renaming with session id; you can do some QC at this point.
    • You can also add nifti files; these are treated as "raw" and "converted" into a format for CSPM analysis, with the ProtocolID in the name and in *.nii format.
  4. From CSPM "Subjects and Study Data", Add subjects.
    • This step can be done earlier; it's independent of the others.
  5. Note which MRI series are to be used. For example, you may want to QC the T1 scans. Some protocols give multiple versions of the scans, e.g., DTI with raw and realigned series, HCP T1 with 4 series for each T1 scan.
  6. Add MRI and physiology studies; you can add comments in the study table (e.g., QC details).
  7. For MRI studies, add path to raw data.
  8. Create project and groups from CSPM "Projects...".
  9. Add project studies, making sure they are linked to subjects.
  10. Add MRI modalities to studies.
  11. Link studies and MRI series to files (as created in step 3); from the Project GUI, Click "Copy" to copy the nifti files from the raw data folder to the project folder. (See details and change from September 2020.) 
  12. Create set of studies and groups to analyze from CSPM "Configure Analysis Sets". All sets should have T1's, so it is best to make sure the "T1" checkbox is checked in the filter section.
  13. Load set and run analysis.
  14. Set preprocessing parameters, run preprocessing.
  15. Check T1's and normalization, and manually align in case of problems (re-run preprocessing if realigned).
  16. Perform statistical analyses.
  17. View results based on SPM or VOI.

Metric Data

Metric data consists of groups of measures. Each group is a "metric." For example, consider the MOCA cognitive test: the MOCA can be classified as a "metric" that includes measures such as the overall MOCA scoree, and the individual component scores (MOCA attention, memory, naming,...etc.).

Metric data are loaded from the Subjects GUI from an Excel file, with one row for each measure, and columns for subject name, metric name, measure type, and measure value. (A template can be created from the Subjects GUI.) These are loaded as "studies" in the Subjects GUI, similar to MRI and physiology data.

Metric studies are configured as part of projects similarly to steps 7-10.

Session ID's

Study ids are a long integer number that uniquely defines the particular study. The convention is:

[subject code number][2-digit session number][8 digit date in international format]

Example: 30130120141004 = subject 3013, session 1 (that is, the first visit or scanning session), Oct 4th 2014.

Related content