MIMESIS Team, Inria, Strasbourg, France.
Université de Strasbourg, CNRS, ICube, Strasbourg, UMR7357, France.
Int J Comput Assist Radiol Surg. 2024 Jun;19(6):1185-1192. doi: 10.1007/s11548-024-03092-4. Epub 2024 Apr 16.
The treatment of cardiovascular diseases requires complex and challenging navigation of a guidewire and catheter. This often leads to lengthy interventions during which the patient and clinician are exposed to X-ray radiation. Deep reinforcement learning approaches have shown promise in learning this task and may be the key to automating catheter navigation during robotized interventions. Yet, existing training methods show limited capabilities at generalizing to unseen vascular anatomies, requiring to be retrained each time the geometry changes.
In this paper, we propose a zero-shot learning strategy for three-dimensional autonomous endovascular navigation. Using a very small training set of branching patterns, our reinforcement learning algorithm is able to learn a control that can then be applied to unseen vascular anatomies without retraining.
We demonstrate our method on 4 different vascular systems, with an average success rate of 95% at reaching random targets on these anatomies. Our strategy is also computationally efficient, allowing the training of our controller to be performed in only 2 h.
Our training method proved its ability to navigate unseen geometries with different characteristics, thanks to a nearly shape-invariant observation space.
心血管疾病的治疗需要对导丝和导管进行复杂且具有挑战性的操控。这通常会导致介入时间延长，使患者和临床医生暴露在 X 射线下。深度强化学习方法在学习这项任务方面显示出了良好的效果，可能是实现机器人介入过程中导管导航自动化的关键。然而，现有的训练方法在泛化到未见的血管解剖结构方面能力有限，每次几何结构发生变化都需要重新训练。
本文提出了一种用于三维自主血管内导航的零样本学习策略。通过使用非常小的分支模式训练集，我们的强化学习算法能够学习到一种控制方法，然后可以将其应用于未见的血管解剖结构，而无需重新训练。
我们在 4 种不同的血管系统上进行了验证，在这些解剖结构上随机目标的平均成功率达到了 95%。我们的策略在计算上也非常高效，仅用 2 小时即可完成控制器的训练。
我们的训练方法通过使用几乎形状不变的观察空间，证明了其在导航具有不同特征的未见几何结构方面的能力。
