An Improved Hybrid Herarchical Model for breast Cancer Detection

Abstract

Breast cancer is believed to be one in all most far reaching causes of death among newlinewomen and second highest reason for deaths among humans. Today millions of newlinewomen are suffering from breast cancer. It is difficult to detect breast cancer in the newlineearly stages due to its dormant nature and very few signs and symptoms. Therefore newlinethe main reason behind the diagnosis of breast cancer is to decrease the death rate newlineby achieving accurate results. Manual screening using mammography is tedious and newlinerequires highly trained experts. Besides huge variability in sensitivity, manual newlinescreening for the identification of disease causing agent is a labor-intensive task. newlineFurther, it is time consuming and depends on patient s stage and requires large newlinenumber of images to be analyzed in one slide. Hence there is a need to automate the newlinediagnostic process to improve the sensitivity and accuracy of the tests. An artificial newlineintelligence based hierarchical fuzzy expert system is developed which consists of newlinerisk parameters, subjective parameters, mammograms and cancerous cell images to newlinediagnose breast cancer with precise results. newlineHierarchical fuzzy expert system, an ease to use interface has been designed which newlinecan help the medical specialists for the early diagnosis of breast cancer. This system newlinecan also be used as a classifier to differentiate the types of breast cancer. The newlinehierarchical system contains two panels risk parameters and subjective parameters newlinebased on the fuzzy rules designed in the system, the system give an accurate result. newlineAfter comparing the fuzzy rules with health care experts, the hierarchical fuzzy newlinesystem was found to be 0.98% accuracy, 0.97% sensitivity and 100% specificity newlinerespectively. Once the patient is diagnosed with cancer, this expert system can help newlinethe doctors to keep track of the patient s health during medication. newlineSecondly the mammograms and cancerous cells positive and negative images for newlineBenign and Malignant recorded under standard image acquisition protocol are newlineconsidered for this work.