Unveiling the Future of Oral Squamous Cell Carcinoma Diagnosis: An Innovative Hybrid AI Approach for Accurate Histopathological Image Analysis

Oral cancer poses a formidable global health threat, demanding urgent attention to combat its devastating impact. Timely detection of oral squamous cell carcinoma (OSCC) is pivotal for successful treatment and improved survival rates. However, manual histopathological analysis, reliant on the expertise of medical practitioners, can be time-consuming and vulnerable to subjective discrepancies. To surmount these challenges and elevate diagnostic outcomes, this research explores the transformative potential of artificial intelligence (AI) in OSCC diagnosis. Three distinct methodologies, namely Gabor + CatBoost, ResNet50 + CatBoost, and Gabor+ ResNet50 + CatBoost, were implemented to leverage the power of AI. By extracting 32 low-level features from the Gabor Filter and 100,532 high-level features from the ResNet50 model, the study adopts principal component analysis (PCA) to mitigate overfitting, retaining the top 4096 components. The extracted features underwent individual classification using CatBoost, followed by concatenation and image classification. Remarkably, the third strategy, which synergized Gabor filtering with ResNet50 feature extraction, along with CatBoost classification, demonstrated the most exceptional performance. Achieving an impressive accuracy of 94.92%, 95.51% precision, 84.30% sensitivity, 95.54% specificity, 94.90% F1 score, and 94.9% AUC, these AI-based approaches herald a new era of accurate and efficient OSCC diagnosis.