Techniques for Classifying Breast Lesions

Classification of breast lesions as malignant or benign from mammograms must be based on information present in the mammogram. In addition, clinical data such as patient age can also be used. A common approach to this task is to extract information (called features) from mammograms and then use a mathematical or statistical model (called a classifier) to make the malignant versus benign assessment. This strategy is not

FIGURE 1 Two 5-cm-square regions of mammograms depicting (a) a malignant mass and (b) a benign mass. Reprinted from [58] with permission.

unique to classification with computers; many radiologists advocate a similar systematic approach in their interpretation of mammograms [66,67]. A typical diagram of a classification technique is shown in Fig. 3.

0 0

Post a comment