Advanced Deep Learning Framework for Cancer Cell Detection using Mammography Images

This research study presents an advanced deep-learning framework aimed at improved cancer cell detection in the contextof mammography images. We used the CBIS-DDSM mammography dataset and a convolutional autoencoder to extract high-dimensional image features. We classified these features using benign and malignant classes based on a Gradient-Boosting Machine (GBM) classifier. The autoencoder was used as a feature extractor to compress the images into a latent space capable of capturing key features without redundancies. Other data preprocessing steps included normalization, resizing, and augmentation to improve model robustness and generalization. The GBM classifier was further finely controlled for its hyperparameters so that it responds effectively to the extracted features and gives proper classifications. To make sure that our method remains practically implementable, we have designed a user-friendly GUI incorporating the AutoEncoder-GBM model. Therefore, the user can easily upload mammography images, preprocess them, and then classify them. Thus, this model will show high accuracy in the detection of cancerous cells while at the same time exhibiting an intuitive interface for its use in a clinical setting by medical professionals.