Automated Template Filling for Abdominal CT Reports Using Language Models

This thesis investigates deep learning methods for the automated generation of structured radiology reports from abdominal CT scans. The work explores how language models can be combined with report templates to produce clinically useful draft reports.