A Deep Learning Framework for breast cancer prediction using Image Processing and Cloud-based Analysis

Women from all over the world are affected by breast cancer as a general problem of health. Early diagnosis and proper detection will go a long way in improving breast cancer survival rates as well as patient outcomes. In this study, we propose a novel deep learning-based framework for the prediction of breast cancer using cloud-based analysis and different image processing techniques. To maximize their potential, the UNet and ResNet50 architectures were combined in the framework for image segmentation and feature extraction respectively. Using a large dataset of breast cancer images, it learns to diagnose correctly as well as locate malignant spots. The system also has real-time predictions and is scalable with efficient cloud-based image processing. The suggested approach has the potential to enhance quick recognition and diagnosis of breast cancer, which could enhance patient outcomes and decrease healthcare expenditures. The findings show that the suggested technique has great levels of accuracy (91.15%), precision (89.94 %), sensitivity (90.84%), F1- score (92 %), and area under the curve (95%).