A Study Of Applying ABC And AIS Algorithm For Breast Tumor Diagnosis Based On Ultrasound Imaging

In recent years, the mortality of breast cancer has gradually increased. The traditional tumor screening is time-consuming and invasive for patients.Traditional puncture procedure such as tumor examination time consuming and is a harmful. Ultrasound examination is one of the common breast tumor detection method, a non-invasive, low price and convenient check method.But ultrasound breast tumor images require experienced physicians to distinguish benign tumors from malignant ones. To improve the classification accuracy and decrease training time of computer-aided diagnosis system for ultrasound breast imaging, we propose approaches that combine feature selection and parameter setting optimization simultaneously based on inteligent optimization algorithms.The major work of this pape includes two parts below:1.we propose a computer aided diagnosis(CAD) system which combines the artificial bee colony(ABC) algorithm and the support vector machine(SVM) to classify ultrasound breast tumor images in this study.This study evaluated 300 ultrasound breast tumor images including300 benign tumors and 120 malignant ones. The breast tumor images first were processed by some image processing methods, such as noise reducing, edge enhancing, and automatic segmentation. The texture features and morphologic features were extracted following the use of an ABC algorithm to detect significant features and determine the optimal parameters for the support vector machine(SVM) to identify the tumor as benign or malignant.The Algorithm’s key is to using artificial colony algorithm of iterative search algorithm, through simulation process of the food source update get optimal solution of the problem.Using ABC algorithm screening characteristics and parameters of SVMThe experimental results show that the accuracy of the proposed CAD system for classifying breast tumors is 94.63%, the sensitivity is 97.78%. Because the proposed CAD system can differentiate benign from malignant breast tumors with high accuracy and short training time, it is therefore to reduce the number of clinically invasive examinations and assist the inexperienced physicians to avoid misdiagnosis.2.Based on feature of artificial immune algorithm that duplicate and selected method,we also combines the artificial immune system(AIS) algorithm with the support vector machine(SVM) to classify ultrasound breast tumor images in this study.The algorithm improves the performance of the ultrasonic image tumor assisted diagnosis system. Before the classification, all the breast tumors were segmented automatically by a level set method. Then, the texture features and shape features were first extracted following the use of an artificial immune system algorithm to detect significant features and determine the near-optimal parameters for the support vector machine to identify the tumor as benign or malignant. The experiment shows that the accuracy of the proposed system for classifying breast tumors is 92-98%, the sensitivity is 97.78%, the specificity is 93.33%, the positive predictive value is 91.67%, the negative predictive value is 98.25%, and it also decreases training time of the support vector machine. It is proved that the use of this system can promote the classification accuracy. Compared with the artificial bee colony algorithm,the average accuracy increased by 2.61% and average time increased by 40.49 seconds due to its higher number of iterations requiredThe study proved that swarm intelligence optimization algorithm can effectively improve the efficiency of tumor image recognition and save time.