Advanced Learning Algorithms with Optimized Pre-processing Techniques for Enhanced Diagnostic Performance in Medical Imaging

This study introduces innovative DCNN-based methods for the early detection and classification of skin lesions, enhancing system accuracy through a combination of architectures, pre-processing techniques, and data augmentation. Multiple DCNN models, including Xception, DenseNet, MobileNet, NASNet Mobile, and Efficient-Nets, were evaluated to test deep learning’s potential in complex, multi-class classification tasks. Training these models on preprocessed datasets with optimized hyperparameters (e.g., batch size, learning rate, dropout) improved classification precision for early-stage skin cancers. Evaluation metrics, such as confusion matrices, accuracy, and loss plots, confirmed high classification efficiency with minimal overfitting, as validation results aligned closely with training. DenseNet-201 and MobileNet-V3 Large demonstrated strong generalization abilities, while EfficientNetV2-B3 and NASNet Mobile achieved the best balance between accuracy and efficiency, making them suitable for practical applications. The application of different augmentation rates per class also enhanced the handling of imbalanced data, resulting in more accurate large-scale detection. In addition, this study explored mammogram image classification into benign, malignant, and normal categories. Comprehensive preprocessing ensured balanced class representation, and EfficientNetV2 models (especiallyB3 and B7) achieved exceptional classification accuracy, attributed to their well-optimized architecture balancing depth, width, and resolution. These models showed high convergence rates and generalization, supporting their suitability for medical imaging tasks using transfer learning.