OCNC: Efficient Learning Mechanism to Identify Breast Cancer Disease using Optimized Convolutional Neural Classification Scheme

Breast cancer is among the most prevalent cancers globally, and early detection is critical for improving survival rates. Ultrasound imaging is a widely used diagnostic tool due to its safety and cost-effectiveness. This study proposes a novel methodology, OCNC: Efficient Learning Mechanism to Identify Breast Cancer Disease using Optimized Convolutional Neural Classification Scheme, to enhance the accuracy and reliability of breast cancer detection. The OCNC model employs a custom-designed Convolutional Neural Network (CNN) architecture optimized for ultrasound image analysis, incorporating techniques such as dropout, batch normalization, and hyperparameter tuning for improved performance. The dataset utilized, sourced from the Breast Ultrasound Images Dataset, underwent comprehensive preprocessing including normalization, noise removal using Wavelet Transform, adaptive contrast enhancement, and data augmentation to ensure high-quality input. The proposed model achieved an impressive accuracy of 98.61%, outperforming nine state-of-the-art models, including ResNet50, InceptionV3, and DenseNet121. The results also demonstrate superior precision (98.45%), recall (98.67%), F1-score (98.56%), and ROC-AUC (99.12%). This work provides a robust and efficient tool for early-stage breast cancer detection, significantly contributing to the field of medical imaging.