肿瘤学中的物理信息深度学习
Physics-Informed Deep Learning in Oncology
AI总结:
本综述整合肿瘤学领域的物理信息深度学习(PIDL)方法,开发适配临床任务的PIDL工作流,并分析其面临的物理模型保真度等关键挑战,为肿瘤学相关研究与应用提供支撑。
摘要:
深度学习已在肿瘤学领域取得显著进展,涵盖癌症检测与诊断、医学图像分析、肿瘤表征、治疗规划、疗效评估及预后预测等方面。然而,传统数据驱动模型往往需要大量标注数据集,在异质性临床场景中的泛化能力有限,且缺乏将癌症发生发展及治疗背后的物理、生理和生物学原理明确融入的机制。物理信息深度学习(Physics-Informed Deep Learning, PIDL)将控制物理定律、机理模型、生理约束及领域知识整合入深度学习框架,为从稀疏、噪声及异质性肿瘤学数据中学习提供了一种有原则的方法。本综述对肿瘤学领域的PIDL方法进行了面向任务的综合,涵盖神经网络、柯尔莫哥洛夫-阿诺尔德网络(Kolmogorov-Arnold networks)、卷积与视觉架构、神经算子、生成模型以及混合物理-数据驱动框架。其应用覆盖癌症检测与诊断、图像重建与分析、肿瘤表征、定量成像、肿瘤进展、治疗反应、放射治疗、治疗规划及多模态临床建模。基于综合的方法学和应用层面证据,我们开发了任务优化的PIDL工作流,使临床任务与合适的数据表示、物理约束、模型架构及验证策略相匹配。我们还分析了关键挑战,包括物理模型保真度、计算可扩展性、不确定性量化、领域泛化、标准化基准测试及临床验证。
英文摘要:
Deep learning has enabled substantial advances in oncology, including cancer detection and diagnosis, medical image analysis, tumor characterization, treatment planning, response assessment, and outcome prediction. However, conventional data-driven models often require large annotated datasets, exhibit limited generalization across heterogeneous clinical settings, and lack explicit mechanisms for incorporating the physical, physiological, and biological principles underlying cancer development and treatment. Physics-informed deep learning (PIDL) integrates governing physical laws, mechanistic models, physiological constraints, and domain knowledge into deep learning frameworks, providing a principled approach for learning from sparse, noisy, and heterogeneous oncological data. This review presents a task-oriented synthesis of PIDL methods across oncology, encompassing neural networks, Kolmogorov-Arnold networks, convolutional and vision architectures, neural operators, generative models, and hybrid physics-data-driven frameworks. Their applications are examined across cancer detection and diagnosis, image reconstruction and analysis, tumor characterization, quantitative imaging, tumor progression, treatment response, radiation therapy, treatment planning, and multimodal clinical modeling. Based on the synthesized methodological and application-level evidence, we develop task-optimized PIDL workflows that align the clinical task with appropriate data representations, physical constraints, model architectures, and validation strategies. We further analyze key challenges, including physical-model fidelity, computational scalability, uncertainty quantification, domain generalization, standardized benchmarking, and clinical validation.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。