Medical Report Generation

Medical Report Generation (MRG) is a task focused on training artificial intelligence to automatically generate professional reports based on input medical imaging data. This task aims to assist clinical doctors in making faster and more accurate diagnostic decisions through automated processes, reducing the time and errors associated with manually writing reports. Although deep neural networks and transformer-based architectures are widely used in this field, the performance of pre-trained models often declines in specific medical domains, primarily due to the significant differences between medical language datasets and general datasets from the internet, as well as the uneven distribution of medical data. Recently, multi-modal learning and contrastive learning have shown potential in MRG, but they still face many challenges that require further research and optimization.