Breast cancer has become a major cause of death among women in developed countries (Senapati et al., 2011). As the causes of breast cancer remain unknown, early detection is crucial to reduce the death rate. However, early detection requires accurate and reliable diagnosis (Cheng et al., 2010). A diagnostic tool should distinguish between benign and malignant tumors while producing low false-positive (rate of missing chances) and false-negative (rate of failure) rates.
Mammography is probably the most effective method for breast tumor detection. However, the technique has limitations in cancer detection. For example, due to its low specificity, many unnecessary biopsy operations are performed, increasing the cost, the emotional pressure, and in some cases the risk to the patient (Zainuddin and Ong, 2010).
In this chapter a wavelet network is constructed to classify breast cancer based on various attributes. Hence, a computer-aided system is developed and proposed to provide additional accuracy in the classification of benign and malignant cases of breast tumors. The Wisconsin breast cancer (WBC) data set was obtained by the UCI Machine Learning Repository and was provided by Mangasarian and Wolberg (1990). In this particular case study we are more interested in producing fewer false negatives. Whereas a false positive will result in extra cost for additional clinical tests, a false negative may result in the death of the patient.