Medical Image Processing - Signal Processing Systems - page 344

in the use of medical imaging and, consequently, the volume of medical image
data being produced. It is estimated that 40,000 terabytes of medical image data
were generated in the United States alone in 2009 [ 8 ] . With expanding use of
medical imaging and growing size and resolution of medical images, medical image
processing has evolved into an important subspecialty of image processing and the
field continues to gain prominence and catch the fancy of the scientific community.
This chapter is designed to introduce the reader to the common medical imaging
modalities (i.e., techniques) first. Most medical images result from the interaction
of some form of energy with the human body, governed by the fundamental
principles of physics. Specialized instruments control the release and recording of
the energy before and after the said interaction. Often the recorded energy (i.e.,
sensed signals) need additional processing or go through the process of image
reconstruction to assume a format that is easy to comprehend and interpret. It is
also possible to make further calculations on the reconstructed images to extract
new or quantitative information. As the number of images per exam has grown
overwhelming the traditional two-dimensional (2D), slice-by-slice review process,
newer algorithms for efficient and intuitive visualization of the entire exam (often
a three-dimensional [3D] image) have become equally important. Clearly, each
imaging modality represents a sophisticated confluence of physics, electronics, data
processing, and computing. We focus primarily on data processing and computing
aspects of medical imaging while keeping the discussion of physics and electronics
to their basic minimum. The learning objectives of this chapter are to acquire an
overview of the common algorithms used in the reconstruction, processing, and
visualization of medical images, and to develop an appreciation for the associated
computing considerations.

2 Medical Imaging Modalities

This section provides an overview of common medical imaging modalities. The
overview comprises a brief discussion to the underlying physics and instrumenta-
tion, the reconstruction problem, and the defining characteristics of the modality.
For an in-depth treatment, the reader is referred to many excellent sources cited in
Sect. 11 .

2.1 Radiography

Commonly referred to as x-ray, conventional radiography (or simply radiography) is
the oldest of all medical imaging modalities. Radiography creates a projection image
by collapsing the 3D anatomic information into a 2D image that is the shadow cast
by the body to the x-rays. As x-rays enter the body, they are absorbed differentially
by various types of tissues along their paths. For example, the hard tissues such