标题： 像素完美 MegaMed：生成高分辨率医学图像的百万像素级视觉-语言基础模型 显示英文标题

标题： Pixel Perfect MegaMed: A Megapixel-Scale Vision-Language Foundation Model for Generating High Resolution Medical Images

摘要： 医学图像合成由于临床环境中固有的复杂性和高分辨率细节要求而面临独特的挑战。 传统的生成架构，如生成对抗网络（GANs）或变分自编码器（VAEs），在高分辨率图像生成方面表现出巨大的潜力，但在保留对准确诊断至关重要的细粒度细节方面存在困难。 为了解决这个问题，我们引入了Pixel Perfect MegaMed，这是第一个能够以1024x1024分辨率合成图像的视觉-语言基础模型。 我们的方法部署了一个多尺度Transformer架构，专门设计用于超高清医学图像生成，从而实现了全局解剖上下文和局部图像级细节的保留。 通过利用针对医学术语和成像模式定制的视觉-语言对齐技术，Pixel Perfect MegaMed在前所未有的分辨率水平上弥合了文本描述与视觉表示之间的差距。 我们将该模型应用于CheXpert数据集，并展示了其从文本提示生成临床真实胸片的能力。 除了视觉质量外，这些高分辨率合成图像在下游任务中也证明是有价值的，当用于数据增强时，特别是在数据量较少的情况下，显示出可衡量的性能提升。 我们的代码可通过项目网站获取 - https://tehraninasab.github.io/pixelperfect-megamed. 显示英文摘要

摘要： Medical image synthesis presents unique challenges due to the inherent complexity and high-resolution details required in clinical contexts. Traditional generative architectures such as Generative Adversarial Networks (GANs) or Variational Auto Encoder (VAEs) have shown great promise for high-resolution image generation but struggle with preserving fine-grained details that are key for accurate diagnosis. To address this issue, we introduce Pixel Perfect MegaMed, the first vision-language foundation model to synthesize images at resolutions of 1024x1024. Our method deploys a multi-scale transformer architecture designed specifically for ultra-high resolution medical image generation, enabling the preservation of both global anatomical context and local image-level details. By leveraging vision-language alignment techniques tailored to medical terminology and imaging modalities, Pixel Perfect MegaMed bridges the gap between textual descriptions and visual representations at unprecedented resolution levels. We apply our model to the CheXpert dataset and demonstrate its ability to generate clinically faithful chest X-rays from text prompts. Beyond visual quality, these high-resolution synthetic images prove valuable for downstream tasks such as classification, showing measurable performance gains when used for data augmentation, particularly in low-data regimes. Our code is accessible through the project website - https://tehraninasab.github.io/pixelperfect-megamed.