An Efficient Breast Cancer Analysis Technique by Using a Combination of HOG and Canny Edge Detection Techniques
Ramisha Anjum, Rubaiya Rahman Dipti, Harun Or Rashid, Shamim Ripon · 2021 5th International Conference on Trends in Electronics and Informatics (ICOEI) · 2021
Breast cancer is a common and frequent cancer among women all over the world and one of the major causes of cancer related deaths. Early and timely detection of breast cancer plays a key role for the proper treatment of the disease. Pathological diagnosis of such cancer is a very important part and if applied in an effective manner would save a lot of lives. The diagnosis includes the analysis of histopathological images of the cancer cells and it is a tedious task and requires a certain level of expertise. This paper proposes a machine learning approach to analyse the histopathological images of breast tissues and shows an improved technique for the detection of malignant (cancerous) cells. Feature extraction from images plays the central role in image processing. A combination of Histograms of oriented gradient (HOG) and Canny Edge detection technique is applied for extracting features which are then reduced by applying Principal component analysis. Classification algorithms like Support Vector Machine (SVM), Logistic Regression (LR) and Adaboost are used to train the proposed model. The experiment shows 94% correct detection of malignant or cancerous cases. Different types of comparisons are shown to identify the suitable method that can help pathologists in the process of breast cancer detection.