MR Tissue Characterization Using Iconic Fuzzy Sets
Tissue characterization in MR imaging mostly uses statistical pattern recognition methods, as is demonstrated by the many papers on this topic. Often, these techniques are based on the intrinsic tissue parameters spin-density N(H), spin-lattice relaxation time T1, and spin-spin relaxation time T2 (e.g., Meindl et al., this volume, p. 174). Other parameters like diffusion or flow may be used as well, and some groups report on the modeling of multi-exponential T2 processes which can be employed for tissue characterization (e.g., Staemmler et al., this volume, p. 63). Others use images with varying contrasts, acquired using different pulse sequences and timing parameters (e.g., Alaux and Rinck, this volume, p. 165), which as such form the basis for relaxation time calculations and thus are highly correlated with the latter.
KeywordsImage Point Tissue Characterization Tissue Parameter Fuzzy Cluster Technique Unsupervised Pattern Recognition
Unable to display preview. Download preview PDF.