[论文解读] Intentional deep overfit learning (IDOL): A novel deep learning strategy for adaptive radiation therapy

Jaehee Chun|arXiv (Cornell University)|Jan 1, 2022

Advanced Radiotherapy Techniques参考文献 20被引用 32

一句话总结

本文提出了一种名为意图性深度过拟合学习（IDOL）的新颖深度学习框架，通过有意识地将模型过拟合至患者特异性先验数据，以提升自适应放疗（ART）中的性能。该方法首先在多样化数据集上训练一个通用模型，随后在单个患者的增强先验信息上进行微调，实现了显著的性能提升——Dice分数提高0.088，MRI超分辨率的平均绝对误差（MAE）降低40%，合成CT的MAE降低65%，证明了其在个性化、数据高效自适应放疗应用中的潜力。

ABSTRACT

Purpose: Applications of deep learning (DL) are essential to realizing an effective adaptive radiotherapy (ART) workflow. Despite the promise demonstrated by DL approaches in several critical ART tasks, there remain unsolved challenges to achieve satisfactory generalizability of a trained model in a clinical setting. Foremost among these is the difficulty of collecting a task-specific training dataset with high-quality, consistent annotations for supervised learning applications. In this study, we propose a tailored DL framework for patient-specific performance that leverages the behavior of a model intentionally overfitted to a patient-specific training dataset augmented from the prior information available in an ART workflow-an approach we term Intentional Deep Overfit Learning (IDOL). Methods: Implementing the IDOL framework in any task in radiotherapy consists of two training stages: (1) training a generalized model with a diverse training dataset of N patients, just as in the conventional DL approach, and (2) intentionally overfitting this general model to a small training dataset-specific the patient of interest (N+1) generated through perturbations and augmentations of the available task- and patient-specific prior information to establish a personalized IDOL model. The IDOL framework itself is task-agnostic and is, thus, widely applicable to many components of the ART workflow, three of which we use as a proof of concept here: the autocontouring task on replanning CTs for traditional ART, the MRI super-resolution (SR) task for MRI-guided ART, and the synthetic CT (sCT) reconstruction task for MRI-only ART. Results: In the replanning CT autocontouring task, the accuracy measured by the Dice similarity coefficient improves from 0.847 with the general model to 0.935 by adopting the IDOL model. In the case of MRI SR, the mean absolute error (MAE) is improved by 40% using the IDOL framework over the conventional model. Finally, in the sCT reconstruction task, the MAE is reduced from 68 to 22 HU by utilizing the IDOL framework. Conclusions: In this study, we propose a novel IDOL framework for ART and demonstrate its feasibility using three ART tasks. We expect the IDOL framework to be especially useful in creating personally tailored models in situations with limited availability of training data but existing prior information, which is usually true in the medical setting in general and is especially true in ART.

研究动机与目标

• 为解决自适应放疗（ART）中有限且高质量的训练数据挑战，利用患者特异性先验信息。
• 克服传统深度学习方法过度强调泛化性而忽视患者特异性性能的局限性。
• 提出一种医学人工智能的新范式，将过拟合有意识地用作个性化精准医疗中的工具。
• 在三个关键ART任务中验证IDOL框架的性能：自动勾画、MRI超分辨率和合成CT生成。

提出的方法

• IDOL框架包含两个训练阶段：首先在包含N名患者的多样化数据集上训练一个通用模型。
• 其次，有意识地将该通用模型过拟合至由患者先期影像和勾画数据经扰动与增强生成的小型患者特异性数据集。
• 该框架与任务无关，可应用于放疗中任意图像到图像的转换或分割任务。
• 通过生成增强的训练样本，利用现有先验知识（如计划CT、勾画或诊断扫描）来提升单个患者的模型保真度。
• 方法采用标准深度学习架构（如U-Net变体），并使用针对任务特定指标（如Dice或MAE）优化的损失函数。
• 该方法避免为每项任务进行超参数调优，转而依赖模型通过受控过拟合记忆并利用患者特异性模式的能力。

实验结果

研究问题

• RQ1 有意识地将模型过拟合至患者特异性先验数据，是否能显著提升深度学习模型在自适应放疗任务中的性能？
• RQ2 在临床放疗中常见的低数据环境下，IDOL框架是否优于传统以泛化为重点的深度学习模型？
• RQ3 先期影像和勾画数据在多大程度上可通过增强手段生成有效的患者特异性训练集？
• RQ4 IDOL框架是否可在包括自动勾画、MRI超分辨率和合成CT生成在内的多种ART任务中实现泛化？
• RQ5 IDOL是否能在无需任务特异性超参数调优的情况下实现高性能，从而具备临床可部署性？

主要发现

• 在重计划CT上的自动勾画任务中，IDOL模型的Dice相似系数（DSC）从0.847 ± 0.031提升至0.935 ± 0.026。
• 在MRI超分辨率任务中，平均绝对误差（MAE）从4.18 ± 0.86降低至2.50 ± 0.54，改善率达40%。
• 在合成CT生成任务中，MAE从68 ± 16 HU降低至22 ± 5 HU，降幅达67%。
• IDOL框架在无需任务特异性超参数调优的情况下实现上述成果，证明其广泛适用性。
• 个性化模型的训练时间极短——自动勾画平均仅20秒，MRI超分辨率为93秒，合成CT生成为3小时。
• 该方法在所有三项任务中均表现出一致的性能提升，验证了其作为ART中患者特异性深度学习策略的可行性。