Breast Cancer Diagnosis from Ultrasonic Image and Histopathology Image Using Deep Learning Approach

Breast cancer is a widespread and potentially life-threatening illness, emphasizing the critical need for early and precise diagnosis to improve treatment outcomes and patient survival rates. In our study, we propose a novel approach to address this challenge by utilizing granular computing, a method that allows for the efficient analysis of complex data by dividing it into smaller, more manageable subsets. Our proposed model harnesses the capabilities of granular computing to analyze breast cancer symptoms extracted from two distinct image modalities: ultrasound images and breast histopathology images. Ultrasound imaging provides real-time, non-invasive visualization of breast tissue, while breast histopathology images offer detailed microscopic views of tissue samples obtained through biopsies. Through rigorous experimentation, we assessed the model's ability to accurately identify breast cancer symptoms and distinguish them from benign conditions. Through rigorous experimentation, our proposed model demonstrated remarkable AUC-ROC rates of 92% in the random forest and 91% in the conventional neural network for ultrasound images and breast histopathology images, respectively. Ultimately, the adoption of such advanced computational techniques has the potential to facilitate timely interventions and improve patient outcomes by enabling clinicians to make more informed decisions based on accurate and reliable diagnostic information. This research contributes to the growing field of medical image analysis by highlighting the potential of granular computing in addressing complex diagnostic challenges such as breast cancer identification.