Абстрактный

Study of Mammogram Microcalcification to aid tumour detection using Naive Bayes Classifier

S.Krishnaveni, R.Bhanumathi, T.Pugazharasan

At present most of the women were having symptoms of breast cancer it can be detected by presence of microclacifications in mammogram. Classifier makes vital role in early detection and diagnosis of microclacifications in mammogram. Most of the classifier at present which is not more efficient to diagnosis the breast cancer. In this paper leads to analysis an efficient method by diagnosing the mammogram using Naive bayes classifier. The proposed method has a) ROI extraction (Chain code) b) Pre-processing (Enhancement), c) Feature extraction (HOG) and d) Classification using Naive bayes classifier. Naive bayes classifier is used to detect microcalcification at each location in the mammogram. It classifies the Mammogram images as Benign or Malignant. The test of the proposed system yield 96.5% microcalcification detection in mammograms. Experimental results show that the proposed method using Mammogram Image Analysis Society (MIAS) Database clinical mammogram.