A Transformer-ResNet-Integrated Deep Learning Model for Digital Pathology-Based Grading and Risk Stratification of MASLD-Related Liver Fibrosis

Key Points

• This research aims to develop a deep learning model for evaluating MASLD-related liver fibrosis from digital pathology images.
• Developed a Transformer-ResNet-integrated model for analyzing liver biopsy histopathology images.
• Utilized digital pathology for automated assessment and risk stratification of liver fibrosis.
• Identified the need for integrating clinical risk factors for better validation.
• The model provides an interpretable assessment of fibrosis staging in liver biopsy samples.
• Offers a quantitative tool for risk stratification based on pathology phenotypes.
• Highlights the necessity for clinical integration and longitudinal validation.