Jenkins Alexander, Osorio Eliana Vasquez, Green Andrew, van Herk Marcel, Sperrin Matthew, McWilliam Alan
Department of Electrical and Electronic Engineering, Imperial College London, London, UK.
Division of Cancer Sciences, University of Manchester, Manchester, UK.
Med Phys. 2025 Jul;52(7):e17919. doi: 10.1002/mp.17919. Epub 2025 Jun 2.
Radiotherapy, the use of high-energy radiation to treat cancer, presents a challenge in determining treatment outcome relationships due to its complex nature. These challenges include its continuous, spatial, high-dimensional, multi-collinear treatment, and personalized nature, which introduces confounding bias.
Existing voxel based estimators may lead to biased estimates as they do not use a causal inference framework. We propose a novel estimator using sparsity via Adaptive Lasso within Pearl's causal framework, the Causal Adaptive Lasso (CAL).
First, simplified 2-dimensional treatment plans were simulated on and grids. Each simulation had an organ at risk placed in a consistent location where dose was minimized and a randomly placed target volume where dose was maximized. Treatment uncertainties were simulated to emulated a fractionated delivery. A directed acyclic graph was devised which captured the causal relationship between our outcome, including confounding. The estimand was set to the associated dose-outcome response for each simulated delivery ( ). We compared our proposed estimator the CAL against established voxel based regression estimators using planned and delivered simulated doses. Three variations on the causal inference-based estimators were implemented: causal regression without sparsity, CAL, and pixel-wise CAL. Variables were chosen based on Pearl's Back-Door Criterion. Model performance was evaluated using Mean Squared Error (MSE) and assessing bias of the recovered estimand.
CAL is tested on simulated radiotherapy treatment outcome data with a spatially embedded dose response function. All tested CAL estimators outperformed voxel-based estimators, resulting in significantly lower total MSE, , and bias, yielding up to a four order of magnitude improvement in compared to current voxel-based estimators ( compared to ). CAL also showed minimal bias in pixels with no dose response.
This work shows that leveraging sparse causal inference methods can benefit both the identification of regions of given dose-response and the estimation of treatment effects. Causal inference methodologies provide a powerful approach to account for limitations in voxel-based analysis. Adapting causal inference methodologies to the analysis of clinical radiotherapy treatment-outcome data could lead to new and impactful insights on the causes of treatment complications.
放射疗法，即使用高能辐射治疗癌症，由于其性质复杂，在确定治疗结果关系方面面临挑战。这些挑战包括其连续、空间、高维、多共线性治疗以及个性化性质，这会引入混杂偏差。
现有的基于体素的估计器可能会导致有偏差的估计，因为它们没有使用因果推断框架。我们在Pearl的因果框架内提出了一种通过自适应套索使用稀疏性的新型估计器，即因果自适应套索（CAL）。
首先，在[具体内容缺失]和[具体内容缺失]网格上模拟简化的二维治疗计划。每次模拟都有一个危及器官放置在剂量最小化的一致位置，以及一个随机放置的靶体积，剂量在该体积处最大化。模拟治疗不确定性以模拟分次给药。设计了一个有向无环图，该图捕获了我们的结果之间的因果关系，包括混杂因素。将估计量设置为每次模拟给药的相关剂量 - 结果响应（[具体内容缺失]）。我们将我们提出的估计器CAL与使用计划和交付的模拟剂量的既定基于体素的回归估计器进行比较。实施了基于因果推断的估计器的三种变体：无稀疏性的因果回归、CAL和逐像素CAL。根据Pearl的后门标准选择变量。使用均方误差（MSE）评估模型性能并评估恢复的估计量的偏差。
在具有空间嵌入剂量响应函数的模拟放射治疗结果数据上测试了CAL。所有测试的CAL估计器均优于基于体素的估计器，导致总MSE、[具体内容缺失]和偏差显著降低，与当前基于体素的估计器相比，[具体内容缺失]提高了多达四个数量级（[具体内容缺失]与[具体内容缺失]相比）。CAL在没有剂量响应的像素中也显示出最小的偏差。
这项工作表明，利用稀疏因果推断方法既有利于识别给定剂量 - 响应区域，也有利于估计治疗效果。因果推断方法为解决基于体素分析的局限性提供了一种强大的方法。将因果推断方法应用于临床放射治疗结果数据分析可能会对治疗并发症的原因产生新的、有影响力的见解。
