用扩散模型预测帕金森病脑部变化，结合用药信息提升准确性。
Treatment-Conditioned Diffusion for Forecasting Neurodegenerative Disease Progression
- 基于Transformer编码器建模用药动态，条件生成未来脑影像。
- 在关键区域加权优化，使生成图像更接近真实解剖结构。
- 相比基线模型，误差降低14%，相似度提升4.9%，适合临床研究使用。
预测神经退行性疾病（如帕金森病）的进展对长期规划和个性化治疗至关重要。现有系统多输出单一临床评分，忽视纵向神经影像的丰富结构；传统生成方法则易丢失解剖细节并模糊细微进展模式。为此，本文提出一种新型治疗条件扩散框架，通过患者基线DaTscan影像与一年内左旋多巴等效日剂量作为条件，预测高保真未来脑状态。该流程采用基于Transformer的编码器表征非线性、时变的药理动态，并通过多权重感兴趣区域掩码聚焦生物关键区域以优化生成。实验表明，本方法保持了清晰的解剖边界，临床保真度显著优于基线，实现14.0%更低的均方误差（MSE）、7.2%更低的平均绝对误差（MAE）及4.9%更高的结构相似性指数（SSIM）。
原文摘要 · Abstract (English)
Forecasting the progression of neurodegenerative diseases, such as Parkinson's disease, is essential for effective long-term planning and personalized therapeutic intervention. Existing systems typically produce scalar clinical scores that ignore the rich structure of longitudinal neuroimaging, while traditional generative approaches suffer from a loss of anatomical details and blurring subtle progression patterns. To address this, we introduce a novel treatment-conditioned diffusion framework that predicts high-fidelity future brain states by conditioning the generative process on patients' screening DaTscan images and levodopa equivalent daily dose over one year. The pipeline uses a Transformer-based encoder to represent non-linear, time-dependent pharmacological dynamics and optimizes generation through a multi-weight region-of-interest mask that focuses on biologically critical areas. Experimental evaluation shows that our framework maintains sharp anatomical boundaries and significantly improves clinical fidelity relative to the baseline, achieving 14.0% lower MSE, 7.2% lower MAE, and 4.9% higher SSIM.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。