Department of Radiation Oncology (Maastro), GROW Research Institute for Oncology and Reproduction, Maastricht University Medical Centre+, Maastricht, The Netherlands.
Medical Image Analysis group, Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, The Netherlands.
Phys Med Biol. 2024 Aug 9;69(16). doi: 10.1088/1361-6560/ad69f6.
The aim of this work was to develop a novel artificial intelligence-assisteddosimetry method using time-resolved (TR) dose verification data to improve quality of external beam radiotherapy.. Although threshold classification methods are commonly used in error classification, they may lead to missing errors due to the loss of information resulting from the compression of multi-dimensional electronic portal imaging device (EPID) data into one or a few numbers. Recent research has investigated the classification of errors on time-integrated (TI)EPID images, with convolutional neural networks showing promise. However, it has been observed previously that TI approaches may cancel out the error presence on-maps during dynamic treatments. To address this limitation, simulated TR-maps for each volumetric modulated arc radiotherapy angle were used to detect treatment errors caused by complex patient geometries and beam arrangements. Typically, such images can be interpreted as a set of segments where only set class labels are provided. Inspired by recent weakly supervised approaches on histopathology images, we implemented a transformer based multiple instance learning approach and utilized transfer learning from TI to TR-maps.. The proposed algorithm performed well on classification of error type and error magnitude. The accuracy in the test set was up to 0.94 and 0.81 for 11 (error type) and 22 (error magnitude) classes of treatment errors, respectively.. TR dose distributions can enhance treatment delivery decision-making, however manual data analysis is nearly impossible due to the complexity and quantity of this data. Our proposed model efficiently handles data complexity, substantially improving treatment error classification compared to models that leverage TI data.
本工作旨在开发一种新的人工智能辅助剂量验证方法，使用时间分辨（TR）剂量验证数据来提高外束放射治疗的质量。虽然阈值分类方法通常用于误差分类，但由于将多维电子射野影像装置（EPID）数据压缩为一个或几个数字会导致信息丢失，可能会导致错过一些误差。最近的研究已经调查了时间积分（TI）EPID 图像上的误差分类，卷积神经网络显示出了希望。然而，之前已经观察到 TI 方法可能会在动态治疗期间消除地图上的误差存在。为了解决这个限制，我们为每个容积调强弧形放射治疗角度模拟了 TR 地图，以检测由复杂的患者几何形状和射束排列引起的治疗误差。通常，这些图像可以被解释为一组仅提供集合类标签的段。受最近对组织病理学图像的弱监督方法的启发，我们实现了一种基于变压器的多实例学习方法，并利用从 TI 到 TR 地图的迁移学习。该算法在错误类型和错误幅度的分类上表现良好。在测试集中，11 种（错误类型）和 22 种（错误幅度）类别的治疗误差的准确率分别高达 0.94 和 0.81。TR 剂量分布可以增强治疗交付决策，但是由于数据的复杂性和数量，手动数据分析几乎是不可能的。与利用 TI 数据的模型相比，我们提出的模型有效地处理了数据复杂性，大大提高了治疗误差分类的效率。
