Head and Neck Tumor Histopathological Image Representation with Pr with Pre- Trained Conv ained Convolutional Neur olutional Neural Network and Vision al Network and Vision Transformer

Image representation via machine learning is an approach to quantitatively represent histopathological images of head and neck tumors for future applications artificial intelligence-assisted pathological diagnosis systems. Objective: This study compares image representations produced by a pre-trained convolutional neural network (VGG16) those vision transformer (ViT-L/14) in terms the classification performance tumors. Methods: Whole-slide five oral tumor categories (n = 319 cases) were analyzed. patches created from manually annotated regions at 4096, 2048, 1024 pixels rescaled 256 pixels. classified logistic regression or multiclass Support Vector Machines binary classifications, respectively. Results: VGG16 with performed best benign malignant salivary gland (BSGT MSGT) (F1 0.703 0.803). outperformed ViT BSGT MSGT all magnification levels. However, maxillofacial bone (MBTs), odontogenic cysts (OCs), (OTs) levels 0.780; 0.874; 0.751). Conclusion: Being more texture-biased, performs better representing