Abstract #2124

Improving the Specificity of Breast MRI with Fuzzy Cluster Analysis

Eric Van Uytven¹, Iain Kirkpatrick², Lawrence Ryner¹

¹National Research Council Institute for Biodiagnostics, Winnipeg, Manitoba, Canada; ²Department of Radiology, University of Manitoba, Winnipeg, Manitoba, Canada

We demonstrate the use of fuzzy cluster analysis (FCA) to analyze dynamic, contrast-enhanced image sets obtained from clinical patients. The analysis highlights the ability of this technique to screen for contra-lateral involvement and reveal complex morphology. This technique also allows the interpreting radiologist to properly analyze tumors with complex uptake kinetics without the need to select a region-of-interest. In summary, complex uptake kinetics and morphological features can be quickly and reliably analyzed, summarized, and presented to the radiologist in a manner which allows for rapid evaluation.