Biomedical Engineering Reference
In-Depth Information
transforms makes them a suitable tool for several applications in signal and
image processing that can benefit from the following advantages:
1. A wavelet transform decomposes a signal to a hierarchy of subbands
with sequential decrease in resolution. Such expansions are especially
useful when a multiresolution representation is needed. Some image seg-
mentation and registration techniques can benefit from a “coarse to fine”
paradigm based on a multiresolution framework.
2. A signal can be analyzed with a multiresolution framework into a spatial-
frequency representation. By carefully selecting the wavelet function and
the space-frequency plane tiling of the transform, distinct components
from a noisy observation signal can be easily separated based on their
spatial-frequency characteristics.
3. Many important features from an image data can be characterized more
efficiently in the spatial-frequency domain. Such feature characterization
was shown to be extremely useful in many applications including registra-
tion and data compression.
In this chapter we summarized some important applications in medical image
processing using wavelet transforms. Noise reduction and enhancement can be
easily implemented by combining some very simple linear thresholding tech-
niques with wavelet expansion. Efficient denoising and enhancement improve
image quality for further analysis including segmentation and registration.
Feature characteristics in wavelet domain were proven to be potentially
more efficient and reliable when compared to spatial analysis only, and therefore
provided more effective segmentation and registration algorithms. We point out
that many other important applications of multiresolution wavelet transforms,
which are beyond the scope of this topic, have not been covered in this chap-
ter, especially image compression, which is considered as one of the greatest
achievements of wavelet transform in recent years [110]. Other important appli-
cations include tomographic image reconstruction, analysis of functional MRI
images, and data encoding for MRI acquisition.
Despite the great success of multiresolutions wavelet transform in medical
imaging applications for the past 20 years, it continues to be a very active area
of research. We list a few resources below that are of interest to readers willing
to acquire more knowledge in research and applications in this area.
