Deep Learning for OSCC Diagnosis: A Multimodal Survey of Techniques, Challenges, and Future Directions
Vinaya R. Kudatarkar, A S Patil, Savita K. Shetty
Abstract: Oral cancer, particularly Oral Squamous Cell Carcinoma (OSCC), remains a significant global health concern due to high mortality rates and frequent late-stage diagnosis. Often identified at an advanced stage because of publicignoranceand restrictions in traditional diagnostic techniques, Oral Squamous Cell Carcinoma(OSCC) is among the most common and lethal types of oral cancer. By providing robust tools for automatic and reliable illness identification, artificial intelligence (AI), especially deep learning, has transformed medical picture analysis in recent years. This survey study offers a thorough assessment of state-of-the-art deep learning techniques used to Oral cancer detection across several imaging modalities including histopathology, fluorescence, hyperspectral, and white light pictures. We methodically investigate and contrast hybrid systems, transformer architectures, transfer learning models, and convolutional neural networks (CNNs) with respect to classification accuracy, resilience, and clinical relevance. The research also addresses real-time deployment issues, model interpretability, and multimodal data integration's importance. Moreover, this study points out present research voids—such as restricted generalizability and absence of stage-wise lesion classification—and offers future research paths to close these obstacles. This effort intends to lead academics, doctors, and developers toward the building of efficient, scalable, and accessible AI-driven diagnostic tools for early OSCC diagnosis and intervention by synthesizing ideas from previous advancements.