Breast cancer diagnosis by choosing the optimal featuresof the fine needle aspiration test using genetic algorithm

Breast cancer is the second leading cause of cancer deaths worldwide and occurrsin one out of eight women. The Fine Needle Aspiration test (FNA) is a low-cost,easy and fast method for accurate and early diagnosis of breast cancer. In thementioned method, the fluid extracted from the breast tissue is tested to check thecytology characteristics so that it can be used to determine whether the mass isbenign or malignant. In cases where it is not possible to determine with certaintywhether the disease is benign or malignant, the use of intelligent computeralgorithms can help doctors in this field. In this study, the data of the WBCDdatabase available at UCI, which includes ۶۹۹ benign and malignant breast tumorsamples, was used. In this database, each sample has ۹ features. First, using thegenetic algorithm, the optimal features were selected from among these ninefeatures, and then the detection of the type of breast cancer in terms of whether itwas benign or malignant was performed using a multilayer perceptron neuralnetwork. For this purpose, the data were divided into two groups, training andtesting. The neural network was trained using training data, and then theperformance of this network was evaluated using test data. The simulation resultsshowed that the algorithm proposed in this study can correctly diagnose the typeof breast cancer in terms of whether it is benign or malignant with ۹۸.۶%accuracy.

Keywords:

Fine needle aspiration test , Breast cancer , Mulilayer Neural network , GeneticAlgorithm