Contents
Prerequisites Protocols
Initial Goal
A set of magnetic resonance (MR) and computer tomography (CT) images of the extremity specimen (upper/lower leg, upper/lower arm) with clear visual delineation of skin, fat, and muscle layers.
Workflow
Schedule MRI/CT
Preparation
Follow the instructions found here, they include registration marker attachment and placement in an imaging fixture
- Double bag using a sterile technique
- Bring USB to transfer MRI data
- CT data is transferred via the Cleveland Clinic servers
- Transport to MRI/CT
Position/Orientation in MRI/CT
- Anatomical position
- Anterior Up
- Distal end of extremity in first
Acquire Image Sequences
For each segment (upper leg, lower leg, upper arm, and lower arm), the following set of image collection protocols will be performed. All images should be acquired in the same coordinate system to be able to align reconstructed tissue geometries during assembly of the full segment geometry. To accomplish this
- The origin (isocenter) and the axes of the magnet, which is set at the beginning of the session, should not change.
- The specimen should not be moved. It should be noted that a pixel-by-pixel alignment of image sets (co-localization) is not necessary.
- Before, each sequence is acquired, verify that the acquisition properties match the desired settings in the tables below. After each sequence is acquired, inspect the images and compare them against the sample images below to to ensure that it was collected properly. If not, the sequence should be recollected.
Notes for CT
- Clean and use stainless push cart found in robot lab to transport specimen
- Ensure operator chooses correct anatomical position
- Use the MULTIS protocol
- Ask to have images placed on CCF server to be downloaded later
MRI Specifications
Image Acquisition Properties: Download T1 SettingsDownload PD Settings
|
T1 w/o Fat Sup |
T1 w/ Fat Sup |
Proton Density |
Plane |
Axial |
|
|
FS |
No |
Yes |
No |
Matrix (phase) |
389 |
|
|
Matrix (freq.) |
512 |
|
|
No. of slices |
25 |
|
|
No. of Seq/Segment |
4 |
|
|
FOV (mm) |
240 x 202.5 |
|
|
Slice thickness/gap (mm/mm) |
2/1 |
|
|
Flip angle (deg.) |
30 |
|
150 |
TE/TR (ms/ms) |
11/500 |
|
9.7/4000 |
Bandwidth (Hz/pixel) |
238 |
|
222 |
Chemical shift (pixels) |
N/A |
|
|
No. excitations averaged |
1 |
|
|
Turbo Factor |
3 |
|
|
Phase encode axis |
Anterior-posterior |
|
|
Distance factor (%) |
50% |
|
|
Phase oversampling |
0 |
|
|
Slice oversampling |
0 |
|
|
Phase resolution (%) |
90 |
|
|
Phase partial Fourier (8/8 = 1) |
OFF |
|
|
Readout partial Fourier (8/8 = 1) |
OFF |
|
|
Slice partial Fourier (8/8 = 1) |
7/8 |
|
|
X-resolution (mm) |
0.5 |
|
|
Y-resolution (mm) |
0.5 |
|
|
Scan Time / Seq. (min.) |
~3 min |
|
|
CT Specifications
Slice thickness/gap (mm/mm) |
0.6/0 |
Reconstruction Interval (mm) |
0.6 |
Recon Type / Kernel |
B40 - medium - Abdomen window |
Voltage |
140 kVP |
Post Processing of CT/MRI
Objective: convert raw dicom files to nifty (.nii) files. The MRI segments must be stitched into upper leg / lower leg (or arm) as it is moved in between. The CT does not require joining and simply needs converted.
Directory Setup
- The MRI image sequences are currently exported as a series of folders, each folder containing one 25 slice sequence of one modality. The T1 and T1 fat sat. are performed right after another on one segment i.e. upper leg. Next, the lower appendage is imaged. The imaging type name and order will make up the filename. The first half of the folders will be for upper, the second for lower. Unless the tech imaged in the opposite order.
- First, simply run the python script dcm2nii.py to convert each sequence to a nifty file.
- Next, reorganize the .nii files into the following directory, which you must first create.
- SubjectID/MRI/nii_leg/FS/LL
- SubjectID/MRI/nii_leg/FS/UL
- SubjectID/MRI/nii_leg/T1/LL
- SubjectID/MRI/nii_leg/T1/UL
- The arm is completed without moving the specimen so there is no UA or LA (upper lower).
Python
Open the ImageUtility folder from the repository containing
- dcm2nii.py
- joinJifty.py
- mytk.py
- Set four parameters as described within the dcm2nii.py script e.g. subjectID, dir. etc. and Run
- Set directory as stated above
- Set the SubID and dir for joinNifty.py, do not change modality
- Change the segment as described in the script. Run
Output
- Compress the raw images at the folder level
- Upload the zipped raw image file for MRI and CT to midas
- Upload the joined .nii files to midas