Cancer is the one of the leading causes of death, affecting millions of lives throughout the world. In 2020, World Health Organization (WHO) reported around 10 million deaths due to various forms of cancer, posing a serious threat to global health and economy. According to WHO reports, Breast cancer is the most common form of cancer (in terms of newly encountered cases), accounting for nearly 2.26 million cases worldwide in 2020. Breast cancer is the most prevalent form of malignancy in women, showing extensive heterogeneity in clinical as well as molecular level. Early breast cancer -where malignant cells are only contained inside the breast or spread only to nearby axillary lymph nodes, is reported to be curable. However, the advanced stage of breast cancer is regarded as treatable disease, for which the primary aim of the various clinical therapies is to prolong patients’ lives by minimising severe symptoms and reducing treatment-associated toxicity to improve the quality of life.
Development of Computer Aided Diagnostic Model for Identification of Breast Malignanancies
A novel Computer Aided Diagnostic Model has been developed to differentiate benign and malignant breast lesions in mammograms using multiresolution analysis and Schmid Filter Bank, which were not reported earlier. A three level Haar wavelet decomposed image(L1, L2, L3) is obtained for each Region of Interest. In each level Texton based analysis is further investigated through Schmid filter bank. Statistical features and Haralick's Features are obtained from filter response and Gray Level Cooccurence Matrix respectively. Partition Membership Filter is further applied to the feature matrix for feature partitioning. The method shows maximum accuracy of 98.63% and Area under Curve of 0.981 using Random Forest Classifier and ten fold cross validation. Another
Computer Aided Diagnostic Model has been developed using
Maximum Response 8 Filter Bank. to identify breast malignancies . Here Haralick's features from the Gray Level Co-occurence Matrix , histogram based features from Local Binary Pattern and statistical features namely skewness and kurtosis are extracted from each filter response. Genetic Algorithm and Linear Discriminant Analysis has been used for feature selection and feature reduction respectively. Classification is performed using three classifiers namely Naive bayes, Logistic Regression and Linear SVM. The proposed algorithm exhibit an Accuracy of 87.5% and Area under Curve of 0.95 using Logistic Regression classifier.
Design and Development of Transillumination Imaging Setup to Detect Breast Abnormalities
Trans illumination is a phenomenon which involves illumination of a material/sample by passing light through it . It finds great use in medicine right from dentistry to imaging of cancerous lesions. The phenomenon of trans illumination can be utilized to construct a device which can not only detect the sinuses in the breast tissues but also help us to differentiate between abnormalities in the breast masses. The setup will involve developing an optical source. The optical source when projected into the breast will produce a trans-illuminated image .The vasculature along with presence of tumors inside the breast can be easily visualized. A micro computer will be interfaced with the set up which will capture images and store it for further processing . Also the system will be very economical, radiation free, portable and thus can be used any number of times even in the slightest of doubt as opposed to other available imaging systems.