Invasive Ductal Carcinoma Classification Using ResNet18 Model on Transfer Learning Concepts

Breast cancer that begins in the milk ducts and spreads to nearby tissue is known as invasive ductal carcinoma (IDC). Machine learning methods may be used to classify IDC, and transfer learning—which uses pre-trained neural networks like ResNet18—can be a potent strategy for this purpose. IDC, which makes up between 70 and 80 percent. It starts in the milk ducts of the breast and can "metastasize" (spread) to neighbouring tissues or other body regions. IDC frequently manifests as a breast lump or tumour that feels hard, irregular, or otherwise dissimilar from the surrounding breast tissue. Changes in the skin over the breast, nipple alterations (such as nipple inversion or discharge), and breast soreness are other potential indicators and symptoms of IDC. To confirm the diagnosis, further diagnostic procedures like mammography, ultrasound, or biopsy are required because not all breast lumps or alterations are suggestive of IDC. In order to treat invasive ductal carcinoma health disparities in its early stages, the problem must first be diagnosed. Our ResNet18 model was shown to have high classification performance for invasive ductal carcinoma classification with participatory research, with an accuracy rate of more than 80%.