A Computer-Aided Diagnosis System for the Detection and Classification of Breast Cancer
Citations
29 citations
References
563 citations
"A Computer-Aided Diagnosis System f..." refers methods in this paper
...The training data are repeatedly presented to the neural network and weights are adjusted until the MSE is reduced to an acceptable value.(10,15,16) The constructed neural network consists of 6 input neurons for the extracted features....
[...]
526 citations
"A Computer-Aided Diagnosis System f..." refers background or methods in this paper
...The ROI begins as a single pixel, the ‘‘seed,’’ and grows iteratively, aggregating with the pixels that have similar intensity and local contrast.(3,5) In this study, the seed point was selected according to the radiologist’s expertise....
[...]
...1% for regionally advanced and metastatic cancer, respectively.(2,3) Of the many breast cancer lesions, masses and microcalcifications are the most important indicators of malignancy, and their detection is integral for early breast cancer diagnosis....
[...]
...The difficulty in detecting masses is a greater challenge because of its form of natural growth from within the epithelial and connective tissues of the breast, making it similar or obscured to normal breast parenchyma.(3) Mass classification using a combination of support vector machine, radial basis function, kernel and waveletbased feature extraction was achieved by Gorgel et al....
[...]
252 citations
"A Computer-Aided Diagnosis System f..." refers methods or result in this paper
...Furthermore, the Az value obtained for this study is within the specified range of 0.80 to 0.90 for mammography analysis as stated by Zilouchian.13 The SE of the classifier obtained is recorded at an even higher value than that obtained in the works of Campanini et al,17 who scored an SE of 84%....
[...]
...90 for mammography analysis as stated by Zilouchian.(13) The SE of the classifier obtained is recorded at an even higher value than that obtained in the works of Campanini et al,(17) who scored an SE of 84%....
[...]
...The following 6 extracted features were used as the classifier’s inputs: entropy, sum entropy, difference variance, energy, contrast, and dissimilarity.(12,13)...
[...]
136 citations
118 citations
"A Computer-Aided Diagnosis System f..." refers background or methods in this paper
...Regardless, different E-SP tradeoffs can be obtained by changing the classification criterion.(12) The classification performance obtained in this study with an Az value of 0....
[...]
...The following 6 extracted features were used as the classifier’s inputs: entropy, sum entropy, difference variance, energy, contrast, and dissimilarity.(12,13)...
[...]