Fiber Tracking Techniques in Magnetic Resonance Diffusion

Tensor Imaging

Grace MichaelsCSUN, Computer Science

Junior

What is MRDTI ?

• A modern technique for visualization at high resolution of the white matter tracts in the brain.

Two Steps…

MRI scanner, determines the orientation of the tracts at each point by measuring diffusion

rates and identifying the direction of the fastest diffusion.

The information about the diffusion rates is encoded into the

diffusion tensor and various integration techniques are used to reconstruct the path of the fiber.

DT-MRI also promises tofurther our understanding

of abnormalities in a variety of diseases such as

• Stroke • Tumors

• Multiple sclerosis • Schizophrenia

Experiments

• 2D • 3D

Experiments

• 2D

Experiments

• 2D

Experiments

• 2D

How the method works?

• We start from initial concentration of test liquid, corresponding to a drop at some point of the tested media.

How the method works?

• The fact that water molecules are more likely to propagate along the fiber then cross it is reflected in the coefficients of diffusion tensor prescribing different rates of propagation in different directions trough the linear diffusion equation.

Diffusion Tensor

 

where is concentration of test molecules at point

- one step forward in time - one step backward in time

),(),(3

1

3

1

txdx

dT

dx

dtx

dt

d

jij

i j i

),( tx 3x

So, we solve the anisotropic equation:

A curvature-based ridge-tracking algorithm

in 2D

Finding end points

Finding end points

Why is fiber tracking important?

• Brain surgery may cause damage to important fiber bundles. Knowledge of their extension could minimize functional damage to the patient.

• White matter can be visualized using fiber traces for a better understanding of brain anatomy.

• Connectivity between different parts of the brain can be inferred, which is useful for functional and morphological research on the brain.

Future work

• Develop 3D experiments

• Get real data from MRI clinic in Pasadena

• Real data experiments

Acknowledgements

• Dr. Alexander Alekseenko, Math Department, CSUN

• NASA CSUN/JPL PAIR Program

• Dr. Carol Shubin, NASA CSUN/JPL PAIR Program director, Math Department

THANK YOU!

QUESTIONS ???