Information Technology Reference
In-Depth Information
12
Fuzziness in Medical Image Processing:
Representation and Models
Miguel Pagola, Aranzazu Jurio, Daniel Paternain, and Humberto Bustince
12.1
Introduction
This chapter discusses Fuzziness in Medical Image Processing. We start summariz-
ing different techniques of medical image acquisition and their main features. Fuzzy
techniques have been widely used in the processing of all these medical images. Ex-
amining the specialized literature, we discuss the fuzziness present in these images,
including:
•
Fuzziness in pixel information. All medical images have a significant amount of
noise, so the information of every pixel is uncertain.
•
Fuzziness in model representation. Some properties of body organs, as for ex-
ample size, location or shape, vary depending on each person, so it is not possi-
ble to determine an exact model of them.
One the most studied topics is the segmentation of medical images. This process is
used as a previous step when creating a human body model or in the diagnosis of
diseases (tumors, mental illness, ...). In this chapter we present a review of the most
popular fuzzy techniques for medical image segmentation:
•
Fuzzy cluster means
•
Fuzzy connectedness
•
Fuzzy rule based systems
•
Fuzzy measures of uncertainty
Finally, we give to the reader a deeper view of the last technique of this list by pre-
senting an example of the use of ignorance functions to segment ultrasound images.
12.2
Medical Imaging
Technical advances have allowed visualization of structures and their functions in
the living human body. The interpretation of these images plays a prominent role
