Zhao Feng, Wu Yizhou, Hu Mingzhe, Chang Chih-Wei, Liu Ruirui, Qiu Richard, Yang Xiaofeng
Department of Radiation Oncology and Winship Cancer Institute, Emory University, Atlanta, Georgia, USA.
School of Computational Science and Engineering, Georgia Institute of Technology, Atlanta, Georgia, USA.
J Appl Clin Med Phys. 2025 Sep;26(9):e70226. doi: 10.1002/acm2.70226.
Medical imaging is fundamental to digital twin technology, enabling patient-specific virtual models of anatomy and physiology. By integrating high-resolution modalities (Magnetic Resonance Imaging (MRI), Computed Tomography (CT), Positron Emission Tomography (PET), ultrasound) with computational frameworks, recent imaging advances now support real-time simulation, predictive modeling, and earlier disease detection. Such capabilities directly inform individualized treatment planning and contribute to more precise, personalized care. Despite remaining challenges-complex anatomical modeling, multimodal integration, and high computational demands-recent advances in imaging and machine learning have significantly enhanced the accuracy and clinical utility of digital twins. The main contributions of our review are: (1) a system-by-system classification of methodologies; (2) evidence that advanced imaging modalities have improved diagnostic accuracy, treatment effectiveness, and patient outcomes beyond conventional approaches; and (3) identification of remaining technical bottlenecks. We further analyze key technical barriers-such as data scarcity and computational complexity-and outline future directions (e.g., AI-driven data augmentation, real-time model optimization) to unlock digital twins' full potential in precision medicine.
医学成像是数字孪生技术的基础，能够生成针对患者的解剖学和生理学虚拟模型。通过将高分辨率模态（磁共振成像（MRI）、计算机断层扫描（CT）、正电子发射断层扫描（PET）、超声）与计算框架相结合，近期的成像技术进展现在支持实时模拟、预测建模和早期疾病检测。这些能力直接为个性化治疗规划提供信息，并有助于实现更精确、个性化的护理。尽管仍存在挑战——复杂的解剖建模、多模态整合和高计算需求——但成像和机器学习方面的近期进展显著提高了数字孪生的准确性和临床实用性。我们综述的主要贡献包括：（1）对方法进行逐个系统的分类；（2）有证据表明，先进的成像模态比传统方法提高了诊断准确性、治疗效果和患者预后；（3）识别剩余的技术瓶颈。我们进一步分析关键技术障碍——如数据稀缺和计算复杂性——并概述未来方向（如人工智能驱动的数据增强、实时模型优化），以释放数字孪生在精准医学中的全部潜力。
