Automated Radiology Report Generation: A Review of Recent Advances

Phillip Sloan,Philip Clatworthy,Edwin Simpson,Majid Mirmehdi

2024-05-29

Abstract:Increasing demands on medical imaging departments are taking a toll on the radiologist's ability to deliver timely and accurate reports. Recent technological advances in artificial intelligence have demonstrated great potential for automatic radiology report generation (ARRG), sparking an explosion of research. This survey paper conducts a methodological review of contemporary ARRG approaches by way of (i) assessing datasets based on characteristics, such as availability, size, and adoption rate, (ii) examining deep learning training methods, such as contrastive learning and reinforcement learning, (iii) exploring state-of-the-art model architectures, including variations of CNN and transformer models, (iv) outlining techniques integrating clinical knowledge through multimodal inputs and knowledge graphs, and (v) scrutinising current model evaluation techniques, including commonly applied NLP metrics and qualitative clinical reviews. Furthermore, the quantitative results of the reviewed models are analysed, where the top performing models are examined to seek further insights. Finally, potential new directions are highlighted, with the adoption of additional datasets from other radiological modalities and improved evaluation methods predicted as important areas of future development.

Computer Vision and Pattern Recognition

What problem does this paper attempt to address?

The main problems that this paper attempts to solve are the current challenges faced by radiologists in generating radiology reports, including the time pressure of report generation, accuracy issues, and diagnostic delays caused by the shortage of radiologists. With the increase in the demands of medical imaging departments, these challenges have an impact on the ability of radiologists to provide reports in a timely and accurate manner. The paper points out that the Automatic Radiology Report Generation (ARRG) technology has great potential, which can assist radiologists in their work, improve the speed and quality of report generation, and thus alleviate these challenges.

Specifically, the paper explores and solves these problems through the following aspects:
1. **Dataset evaluation**: Evaluate the existing ARRG datasets, including the availability, size, and adoption rate of the datasets.
2. **Deep - learning training methods**: Research on the application of deep - learning training methods such as contrastive learning and reinforcement learning in ARRG.
3. **Model architecture**: Explore the latest model architectures, such as variants of CNN and Transformer models.
4. **Multi - modal input and knowledge graph**: Discuss how to integrate clinical knowledge through multi - modal input and knowledge graph.
5. **Model evaluation techniques**: Review the current model evaluation techniques, including common Natural Language Processing (NLP) metrics and qualitative clinical reviews.

Through this systematic review of these aspects, the paper aims to provide a comprehensive reference for researchers in the ARRG field, promote the development of this field, and propose future research directions, such as using more datasets of other radiology modalities and improving evaluation methods.