c

In this work, a novel classification system for the analysis of mammographic microcalcifications (µCa) using Machine Learning techniques is presented. The automatic classification is performed through Case-Based Reasoning (CBR), which integrates in one system two different characteristics: machine learning capabilities and problem solving capabilities. CBR uses a similar philosophy to that which humans sometimes use: it tries to solve new cases (examples) of a problem by using old previously solved cases. We studied the application of CBR classification to the problem of differentiate benign from malignant µCa in mammograms, obtained from the mammography database of the Girona Health Area, and compared the classification results to other classification techniques