Thermal Breast Cancer Screening via Auto-ROI Asymmetry Segmentation and Classification

Breast cancer is a leading cause of mortality among women, making early detection crucial for survival rates. Thermography, a non-invasive technique, shows promise for early screening by analyzing temperature asymmetries between breast regions. However, precise segmentation of Regions of Interest (ROIs) in thermograms is essential for classification accuracy. This paper presents a fully automated method for segmenting thermograms using a Convolutional-Deconvolutional Neural Network (C-DCNN) and right-left breast segmentation for asymmetry analysis. Followed by examining Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and Random Forest to assess breast health. Experimental results demonstrate competitive segmentation accuracy, with SVM outperforming other classifiers. This study underscores the effectiveness of integrating deep learning with traditional machine learning in breast cancer diagnosis, paving the way for more accurate automated methods.