AI Driven Detection of Skin and Oral Cancer: A Survey of Machine Learning and Deep Learning Approaches

Machine learning (ML) and deep learning (DL) are increasingly important in enhancing the diagnosis of skin and oral cancers, where early detection is critical for improving patient outcomes. Traditionally reliant on visual assessments and clinical tests, cancer diagnosis can be prone to human error. ML and DL offer a scalable, automated approach, analyzing medical images and genomic data to identify patterns that may be missed by humans. Advanced models like Inception-v3 and ResNet have shown promise in improving skin cancer diagnosis, while techniques such as Random Forest and Support Vector Machines (SVM) are used for oral cancer analysis. However, small and imbalanced datasets remain a major challenge, affecting model performance. Key issues like data preprocessing, feature selection, and model training also limit the effectiveness of these methods. Recent advancements, such as explainable AI and multimodal data fusion, are improving the transparency and clinical acceptance of these models. However, a critical issue persists: ensuring that models can efficiently handle real-world data without delays, which is crucial for their integration into patient care. This paper explores these developments and challenges, highlighting the need to address these issues before ML and DL can be fully integrated into clinical cancer diagnosis.