Deep Learning and Optimization Approaches in Semantic Segmentation and Classification for Ovarian Cancer Detection Using EfficientNetB0 with FPN and Causal Dilated Convolutional Neural Networks: A Review

Abstract

Ovarian cancer remains one of the most lethal gynecological malignancies due to its asymptomatic nature in early stages and late diagnosis. Advances in medical imaging and artificial intelligence have enabled automated systems for early detection and classification, improving clinical outcomes. Deep learning techniques, particularly convolutional neural networks (CNNs), have shown strong potential in enhancing diagnostic accuracy through semantic segmentation and classification of histopathological and radiological images. This review highlights deep learning and optimization approaches for ovarian cancer detection, focusing on EfficientNetB0, Feature Pyramid Networks (FPN), and causal dilated convolutional neural networks (CDCNNs). EfficientNetB0 provides efficient feature extraction with reduced computational complexity, making it well-suited for medical imaging tasks, while FPN improves segmentation by integrating multi-scale feature representations for precise tumor localization. CDCNNs enhance contextual understanding by expanding receptive fields without increasing computational cost, enabling better capture of spatial dependencies. The integration of these architectures supports hybrid frameworks capable of accurate segmentation and classification. Despite improved performance, challenges such as limited annotated data, model interpretability, and computational demands persist, encouraging further research in explainable AI, multimodal integration, and lightweight models for real-time clinical deployment.