Web-Based Diagnostic Platform for Breast Cancer Detection Using CNN-GRU

Abstract

Breast cancer is one of the most prevalent and life-threatening diseases affecting women globally. Early and accurate detection plays a crucial role in improving survival rates and enabling timely treatment. With recent advancements in artificial intelligence, deep learning models have emerged as powerful tools for automated disease diagnosis through medical imaging. This paper presents a novel hybrid deep learning approach that integrates Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs), specifically Long Short-Term Memory (LSTM) units, to enhance the accuracy and reliability of breast cancer classification from image data. The CNN component is utilized for extracting spatial features from mammogram images, while the LSTM layer models the sequential and temporal relationships among these features, capturing higher-level patterns that are often overlooked by traditional CNNs. Additionally, the architecture incorporates ReLU activation functions for efficient gradient flow and a Softmax output layer to convert raw prediction scores into interpretable class probabilities. The model was implemented using Python in a Jupyter Notebook environment and trained on a pre-processed breast cancer dataset. Extensive experiments demonstrate that the proposed system achieves a high classification accuracy of 99%, with excellent performance across key metrics such as precision, recall, F1-score, and area under the ROC curve. Furthermore, a web-based interface was developed to facilitate user interaction, allowing real-time prediction by uploading test images. This hybrid model, with its robust design and accessible deployment, represents a promising advancement toward automated, scalable, and clinically applicable breast cancer diagnosis.