The use of transformers, a type of machine learning model, has shown promising results in both natural language processing and vision tasks. These models have the ability to capture global patterns in images and can easily integrate non-image features. In the field of histopathology, digitalization is increasing and AI-based tissue biomarkers are being implemented. Deep learning, particularly suitable for analyzing large images, has shown success in improving the diagnosis of hepatocellular lesions. However, further research is needed to determine its effectiveness in other types of lesions.
