Automatic Medical Report Generation: Methods and Applications
arXiv:2408.13988 · doi:10.1561/116.20240044
Abstract
The increasing demand for medical imaging has surpassed the capacity of available radiologists, leading to diagnostic delays and potential misdiagnoses. Artificial intelligence (AI) techniques, particularly in automatic medical report generation (AMRG), offer a promising solution to this dilemma. This review comprehensively examines AMRG methods from 2021 to 2024. It (i) presents solutions to primary challenges in this field, (ii) explores AMRG applications across various imaging modalities, (iii) introduces publicly available datasets, (iv) outlines evaluation metrics, (v) identifies techniques that significantly enhance model performance, and (vi) discusses unresolved issues and potential future research directions. This paper aims to provide a comprehensive understanding of the existing literature and inspire valuable future research.
42 pages and 9 figures
References in corpus (33)
- Representation Learning with Contrastive Predictive Coding
- The Curious Case of Neural Text Degeneration
- BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models
- BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation
- Align before Fuse: Vision and Language Representation Learning with Momentum Distillation
- COVID-CT-Dataset: A CT Scan Dataset about COVID-19
- PadChest: A large chest x-ray image dataset with multi-label annotated reports
- MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models
- Contrastive Learning of Medical Visual Representations from Paired Images and Text
- MIMIC-CXR-JPG, a large publicly available database of labeled chest radiographs
- Hybrid Retrieval-Generation Reinforced Agent for Medical Image Report Generation
- Knowledge Matters: Radiology Report Generation with General and Specific Knowledge
- Interactive and Explainable Region-guided Radiology Report Generation
- Improving Chest X-Ray Report Generation by Leveraging Warm Starting
- NegBio: a high-performance tool for negation and uncertainty detection in radiology reports
- CXR-CLIP: Toward Large Scale Chest X-ray Language-Image Pre-training
- RadGraph: Extracting Clinical Entities and Relations from Radiology Reports
- Auto-Encoding Knowledge Graph for Unsupervised Medical Report Generation
- CAMANet: Class Activation Map Guided Attention Network for Radiology Report Generation
- Multimodal Image-Text Matching Improves Retrieval-based Chest X-Ray Report Generation
- Cross-modal Memory Networks for Radiology Report Generation
- Competence-based Multimodal Curriculum Learning for Medical Report Generation
- Radiology Report Generation Using Transformers Conditioned with Non-imaging Data
- Improving Radiology Report Generation Systems by Removing Hallucinated References to Non-existent Priors
- PRIOR: Prototype Representation Joint Learning from Medical Images and Reports
- Fine-Grained Image-Text Alignment in Medical Imaging Enables Explainable Cyclic Image-Report Generation
- Effectively Fine-tune to Improve Large Multimodal Models for Radiology Report Generation
- MedCycle: Unpaired Medical Report Generation via Cycle-Consistency
- Unify, Align and Refine: Multi-Level Semantic Alignment for Radiology Report Generation
- PromptMRG: Diagnosis-Driven Prompts for Medical Report Generation
- WsiCaption: Multiple Instance Generation of Pathology Reports for Gigapixel Whole-Slide Images
- Dynamic Traceback Learning for Medical Report Generation
- Sam-Guided Enhanced Fine-Grained Encoding with Mixed Semantic Learning for Medical Image Captioning