9th International Conference on ICT for Sustainable Development

Authors - Sachi Joshi, Upesh Patel
Abstract - The development of particular and effective diagnostic tool is imperative as breast cancers continues to be a major worldwide health problem. Convolutional Neural Networks one form of deep learning approach, have proven wonderful overall performance in medical image evaluation responsibilities in present day years. The method provided in this paper demonstrates CNNs' advantages inside the detection of breast cancers. This paper used the CNN for diagnosing breast cancer using CBIS-DDSM (Curated Breast Imaging Subset of DDSM). The suggested technique first makes use of a CNN architecture to identify patterns that are suggestive of breast abnormalities via extracting excessive level features from mammography images and histopathology images. These features are then extracted and fed into in default MLP classifier, which completes the very last category task and further refines the illustration. By the usage of CNNs' hierarchical feature learning capabilities, this method increases detection accuracy via a synergistic effect. Extensive experiments are achieved using publicly to be had datasets of images so as to assess the efficacy of the technique. The findings show that almost about breast cancers detection task, the CNN version outperforms more conventional machine learning algorithms in phrases of accuracy, sensitivity, and specificity.