Automatically generating a report from a patient's Chest X-Rays (CXRs) is a promising solution to reducing clinical workload and improving patient care. However, current CXR report generators -- which are predominantly encoder-to-decoder models -- lack the diagnostic accuracy to be deployed in a clinical setting. To improve CXR report generation, we investigate warm starting the encoder and decoder with recent open-source computer vision and natural language processing checkpoints, such as the Vision Transformer (ViT) and PubMedBERT. To this end, each checkpoint is evaluated on the MIMIC-CXR and IU X-Ray datasets. Our experimental investigation demonstrates that the Convolutional vision Transformer (CvT) ImageNet-21K and the Distilled Generative Pre-trained Transformer 2 (DistilGPT2) checkpoints are best for warm starting the encoder and decoder, respectively. Compared to the state-of-the-art ($\mathcal{M}^2$ Transformer Progressive), CvT2DistilGPT2 attained an improvement of 8.3\% for CE F-1, 1.8\% for BLEU-4, 1.6\% for ROUGE-L, and 1.0\% for METEOR. The reports generated by CvT2DistilGPT2 have a higher similarity to radiologist reports than previous approaches. This indicates that leveraging warm starting improves CXR report generation. Code and checkpoints for CvT2DistilGPT2 are available at https://github.com/aehrc/cvt2distilgpt2.

翻译：从患者的胸部X光片（CXR）自动生成报告是减少临床工作量并改善患者护理的一种有前景的解决方案。然而，当前的CXR报告生成器（主要为编码器-解码器模型）缺乏在临床环境中部署所需的诊断准确性。为了改进CXR报告生成，我们研究了利用最新的开源计算机视觉和自然语言处理检查点（如Vision Transformer (ViT) 和 PubMedBERT）对编码器和解码器进行暖启动。为此，我们在MIMIC-CXR和IU X-Ray数据集上分别评估了每个检查点。我们的实验研究表明，Convolutional vision Transformer (CvT) ImageNet-21K 和 Distilled Generative Pre-trained Transformer 2 (DistilGPT2) 检查点分别最适合用于编码器和解码器的暖启动。与当前最佳方法（$\mathcal{M}^2$ Transformer Progressive）相比，CvT2DistilGPT2在CE F-1上提升了8.3%，在BLEU-4上提升了1.8%，在ROUGE-L上提升了1.6%，在METEOR上提升了1.0%。CvT2DistilGPT2生成的报告与放射科医生报告的相似度高于以往方法。这表明利用暖启动能有效改进CXR报告生成。CvT2DistilGPT2的代码和检查点可在https://github.com/aehrc/cvt2distilgpt2获取。