First Place Solution of 2023 Global Artificial Intelligence Technology Innovation Competition Track 1
This work addresses the challenge of automated report generation in medical imaging, which could assist radiologists, but it is incremental as it builds on existing models and competition-specific tasks.
The paper tackled the problem of generating medical imaging diagnosis reports by developing a solution that achieved first place in a competition, with a single model scoring 2.321 and fused models scoring up to 2.362 on leaderboards.
In this paper, we present our champion solution to the Global Artificial Intelligence Technology Innovation Competition Track 1: Medical Imaging Diagnosis Report Generation. We select CPT-BASE as our base model for the text generation task. During the pre-training stage, we delete the mask language modeling task of CPT-BASE and instead reconstruct the vocabulary, adopting a span mask strategy and gradually increasing the number of masking ratios to perform the denoising auto-encoder pre-training task. In the fine-tuning stage, we design iterative retrieval augmentation and noise-aware similarity bucket prompt strategies. The retrieval augmentation constructs a mini-knowledge base, enriching the input information of the model, while the similarity bucket further perceives the noise information within the mini-knowledge base, guiding the model to generate higher-quality diagnostic reports based on the similarity prompts. Surprisingly, our single model has achieved a score of 2.321 on leaderboard A, and the multiple model fusion scores are 2.362 and 2.320 on the A and B leaderboards respectively, securing first place in the rankings.