AUTOMATED INHOMOGENEITY CORRECTION

1
AUTOMATED INHOMOGENEITY CORRECTION

• By
• Anuradha Subramanian
• MRI Institute
• Image Retreat 2005

2
What is RF inhomogeneity?

• Inhomogeneity refers to the slowly varying
  intensity patterns that occur in MRI and that
  tend to obfuscate anatomic detail
• The causes
• Non-uniform B1 field
• Non-uniform sensitivity in
• a receive only coil

3
Why is this correction important?

• Inhomogeneity in RF causes unnecessary
  intensity variations in an image
• Causes problems for threshold-based
  segmentation
• Causes problems in the analysis of structural
  information in high field MR images

4
Overview of the correction Algorithm

• The approach is to normalize the input image by
  smoothed version of itself, so that the low
  spatial frequencies are removed.
• The advantage of this method, is that it does
  not cause any prominent artifacts at the edges
  of structures.
• No large artificial increase in MR signal
  intensity in the image contents at the edges of
  objects This is achieved because filling is
  done locally by using the technique of Analytic
  Continuation.

5
Original
Corrected
6
Other Projects

• Filter to remove Gibbs Ringing Artifacts.
• 3 Dimensional Reformatting and Visualization.
• Reformatting of images can be achieved from
  any orthogonal view to any other
  orthogonal view