airvlab/CathAction

CathAction是一个大规模的数据集，旨在推进导管插入理解和碰撞检测的研究。该数据集包含用于导管插入理解和碰撞检测的标注帧，以及专门用于导管和导丝分割的地面实况掩码。数据集分为两个主要部分：导管插入动作理解和碰撞检测。导管插入动作理解部分包含五个类别的动作标注，数据集结构包括视频帧、预提取的RGB特征和训练/验证的CSV文件。碰撞检测部分包含图像和标签文件，用于训练和验证模型。

The CathAction dataset is a large-scale dataset designed for advancing catheterization understanding. It comprises annotated frames focused on catheterization understanding and collision detection, along with groundtruth masks dedicated to catheter and guidewire segmentation. The dataset is divided into two main parts: 1. Catheterization action understanding, including tasks such as catheterization anticipation and action recognition, with five classes: advance catheter, retract catheter, advance guidewire, retract guidewire, and rotate. The dataset includes video frames, pre-extracted RGB features, and CSV files for training and validation. 2. Collision detection, which involves identifying whether the tip of the catheter or guidewire comes into contact with the blood vessel wall. The dataset includes images, label files, and text files for training and validation.