Breast cancer is the most common type of cancer among women worldwide, making its early detection crucial to reduce mortality rates. While mammography is the standard, we explored medical thermography as a non-invasive and low-cost complementary tool. We developed an automated methodology to segment the breast region using vertical histogram projections, eliminating manual intervention. Our approach combines preprocessing techniques and Canny edge detection to extract first-order radiomic features, analyzing thermal asymmetry between breasts of the same patient. Preliminary results show that variance and kurtosis indicate possible thermal heterogeneity. Future work will expand the dataset and implement machine learning algorithms to enhance diagnostic accuracy, aiming for a more automated and robust clinical screening system.