Enhancing radiology report generation through pre-trained language models

Keyphrases

• Encoding Method 100%
• X-ray Images 100%
• Positional Encoding 100%
• Video Encoder 100%
• Pre-trained Language Model 100%
• Radiology Report Generation 100%
• Information Integration 50%
• Medical Records 50%
• Radiologists 50%
• Clinical Patients 50%
• Computational Resources 50%
• Medical Imaging 50%
• Quality Measures 50%
• Multimodal Architecture 50%
• N-gram 50%
• Artificial Intelligence Models 50%
• Healthcare Domain 50%
• Process Data 50%
• Transformer Architecture 50%
• Healthcare Field 50%
• Clinical Notes 50%
• Proposed Architecture 50%
• CheXNet 50%
• Attention Map 50%
• Text Quality 50%
• MIMIC-CXR 50%
• BERTScore 50%
• Standard Text 50%
• Text Summarization 50%
• Model Training 50%
• Pre-trained Large Language Models 50%
• Multimodal Transformer 50%
• Publicly Available Datasets 50%
• Recall-Oriented Understudy for Gisting Evaluation (ROUGE) 50%
• Free-text Radiology Report 50%

Computer Science

• Report Generation 100%
• Positional Encoding 100%
• Pre-Trained Language Models 100%
• Computational Resource 50%
• Artificial Intelligence Model 50%
• Large Language Model 50%
• Integrate Data 50%
• Process Data 50%
• Coder 50%
• Pre-Trained Model 50%
• Multimodal Transformer 50%
• Quality Measure 50%
• Text Summarization 50%