FPGA Implementation for Automated Detection of Breast Cancer using Wavelet Transform

The mortality rate due to breast cancer is increasing, for five minute a women dies worldwide. Mammogram is one which reduces the death rate by early diagnosis and regular screening. This paper hypothesize a consolidate approach of FPGA implementation for automatic detection of breast cancer and classification of the tumor by adopting 2D discrete wavelet transform and artificial neural network (ANN). Initially mammogram is segmented from the background which improves the quality of the image by reducing noise followed by wavelet transform implemented both on software and FPGA hardware. The tools used are MATLAB 2010b and wavelet transform is implemented in Verilog and simulated in Xilinx 13.1 ISE. The paper posits a system which serves as a second opinion to the radiologists and helps in classifying the abnormalities. The images are tested and classified by using Mammography Image Analysis Society (MIAS) database. KeywordsBreast cancer, mammography, Computer Aided Diagnosis (CAD),FCM Clustering, Wavelets, feature extraction, Artificial Neural Network (ANN).