DTI Pipeline

This article is a stub. It needs improvement.

Introduction

The DTI pipeline of MIPAV prepares diffusion weighted images (DWI) and computes voxel-wise diffusion tensors (DT) for the further analysis of diffusion tensor imaging (DTI) data. The pipeline computes maps of diffusion eigenvalues and eigenvectors. It also determines an anatomical correspondence between DTI and structural MRI images of the same sample.

MIPAV DWI pipeline overview

MIPAV DTI Pipeline schematic

1. A user uploads a DWI image and T2 image to the pipeline using the Import Data panel. A DWI image can be acquired from many different MRI scanners (including Philips, Siemens, GE, etc.) and in various formats. MIPAV reads the gradient information from the image header, or from the B-matrix file uploaded by the user. The gradient (or B-matrix) information is then displayed in a gradient table.
2. In the Pre-processing step, the B0 slice in DWI image is detected and it is rigidly registered to the T2 image. The DWI image is then rigidly registered to B0. This steps are necessary to perform a motion correction and eddy current distortion correction.
3. In the EPI Distortion Correction step, MIPAV calculates deformation vector fields for rigidl aligned B0 and T2, which came from the Pre-processing step. MIPAV then uses both: the transformation matrices obtained in the Pre-processing step, and deformation vector field values to create a corrected DWI image.
4. MIPAV then creates a tensor using pre-processing DWI and the gradient/B-value information and a mask image uploaded by the user.
5. MIPAV uses the tensor information to create a whole bunch of images, including ADC, color map, Eigen value, Eigen vector, FA, RA, and Volume Ratio.
6. MIPAV created 3D visualization of fiber bundle tracts in the brain's white matter using the information from the previous step. The user can save fiber tracts information as .vtk and .dat files.

MIPAV DTI Pipeline dialog

Import Data tab

Upload DWI Image

DTI Pipeline reads raw data, all MIPAV supported formats and DICOM (Digital Imaging and Communications in Medicine) formatted files. DWI Image Browse – check this option if you would like to upload your image of interest from your computer. Use Active DWI Image – check this option to use an active image, which is already opened in MIPAV. Note: in MIPAV, the active image is the one that has a red frame.

Using a Structural Image as Reference Space

Brain MRI should be performed with a 1.5- or 3-T MRI machine. High gradient strength in the range of 20-60 mT/m with a slew rate of 120 T/m/s is ideal. Typical parameters for a single-shot spin-echo echo-planar imaging (EPI) sequence are a repetition time (TR) of 6000 ms, an echo time (TE) of 100 ms, and a field of view of 24 cm to obtain 3- to 5-mm axial or coronal sections with a 5-mm intersection gap. The acquisition matrix is 96 × 96 with a reconstruction matrix of 128 × 128. The images are obtained by using 4 linearly increasing b values in 6-7 noncolinear directions (bmax = 703-1000 s/mm2). In addition, a T2-weighted (T2W) image is obtained without diffusion weighting (b = 0 s/mm2).

Why do we need T2 image as a reference space?

In an image-based registration scheme, one uses a cost function Q to measure how well the images are spatially aligned. First, a target image is chosen as a reference for all other images in the data set (source images). Because it is usually less distorted and has a higher signal-to-noise ratio (SNR) than the heavily DW images, the image acquired with no diffusion sensitization (the T2-weighted image), is usually used as the target image for registering DW images. Next, using a spatial transformation model, one aligns all other images to the target image by optimizing a cost function. Image-based registration schemes differ from each other in terms of 1) the definition of Q, 2) the types of transformations applied to the image in searching for the maximum of Q, and 3) the numerical optimization method used in searching for the maximum of Q. We chose to work with an image-based registration scheme because a field map, which is usually not available for DW images acquired on conventional clinical scanner, is not required. In addition, an image-based scheme allows one to correct for mis-registration produced by subject motion.

DTI_base: used as moving image to compute the registration with a T2 reference DTI:final registration transform will be applied to the tensor to resample it in the new reference space (T1). DWI DTI_mask: the mask will be used to guide the automated intensity-based registration of the DTI_baseline. Particularly the nonrigid aspects of the registration to correct for the DTI distortions benefit from the ROI provided by the mask

Upload B-Value/Gradient File or B-Matrix file

B-value A diffusion gradient can be represented as a 3D vector q. The direction of vector q is in the direction of diffusion and its length is proportional to the gradient strength. The gradient strength, or more often used diffusion weighting parameter, is s expressed in terms of the b value parameter, which is proportional to the product of the square of the gradient strength q and the diffusion time interval (b ~ q2 • Δ).

Gradient or B-Matrix table

In the Gradient or B-Matrix table the B column contains b values and X Gradient, Y Gradient And Z gradient columns contain the diffusion gradients applied along the x (Gx), y (G,), and z (Gz) axis. Gradient Creator Input Parameters • Fatshift • Jones 30 • Kirby • Gradient Resolutions • Gradient Overplus • Phillips release • OS • Inverted

Re-Ordering data (from Camino)

Probably the most common task is to re-order data from image to voxel order. In scanner order or image order, multi-component images are stored as consecutive volumes. This is convenient for visualization, since you can easily render a particular 3D volume. It is inconvenient for parallel processing, as you must read the entire 4D image in order to get the components for a particular processing. Data in voxel order stores all components for a particular voxel together. Thus you can read the image one voxel at a time, or skip ahead to particular voxel, without reading the entire image into memory. Camino does most of its I/O in voxel order. You can get into and out of scanner order with the scanner2voxel and voxel2scanner commands. Since these commands expect to deal with mostly raw data, they read and write floats by default (see below). You can change this behaviour with -inputdatatype and -outputdatatype options. For example:

scanner2voxel -voxels 983040 -components 60 -inputfile ScannerOrder.img -inputdatatype short > VoxelOrder.Bfloat where -components specifies the number of volumes in the 4D input and -voxels specifies the number of voxels (ie, X×Y×Z, where X, Y, Z are the dimensions of the image).