Answers are grounded in this paper and cite their sources.
One of the advantages of Magnetic Resonance images is their ability to discriminate tissues for their subsequent quantification. In this work, magnetic resonance brain images are analyzed pixelwise by fuzzy logical predicates, reproducing in a computational way the considerations that experts employ when they interpret these images, in order to identify the tissues that pixels represent. We used Compensatory Fuzzy Logic operators to implement the logical connectives. The problem has been addressed as one pertaining to the discipline of decision-making support. The aim is to determine which tissue corresponds to each pixel. The system is optimized by a Genetic Algorithm that search an adequate set of parameters for fuzzy sets included in the predicates. It has been possible to successfully discriminate cerebrospinal fluid, gray matter and white matter in simulated and real images, validating the results using the Tanimoto Coefficient. As the operations involved are simple, processing time is short. The method can be expanded and adapted to be applied on other types of images and to recognize a greater number of tissues.
No takes yet. Share an insight, caveat, or question.
Meschino et al. (2008) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: