基于AI视觉的手背静脉注射机器人持针算法

Needle-Holding Algorithm for AI Vision Based Hand Vein Injection Robot

摘要: 静脉注射作为一种基础的医疗操作，在临床治疗领域得到了广泛的应用，通常被视为治疗过程中的关键步骤。为减轻医护人员的工作负担，本研究提出了一种机器人辅助静脉注射设备，该设备能够自动执行针头穿刺手背血管的操作。本文简要阐述了该设备的核心组件及其操作界面设计。设备的核心技术在于自动识别穿刺点和辅助穿刺点的能力。研究中采用了YOLOv8算法对数据集进行训练，以获得相应的识别模型，并对其进行了验证测试。此外，本文还对识别到的穿刺点如何转换为机器人末端执行器的坐标进行了初步探讨。

Abstract: As a fundamental medical procedure, intravenous injection has been widely applied in clinical treatment and is generally regarded as a critical step in therapeutic processes. To alleviate the workload of medical staff, this study proposes a robot-assisted intravenous injection device capable of automatically performing needle insertion into the dorsal hand veins. This paper briefly elaborates on the core components of the device and its operational interface design. The key technology lies in the device’s ability to automatically identify puncture points and assist in needle insertion. The YOLOv8 algorithm was employed to train the dataset for obtaining the corresponding recognition model, followed by validation testing. Furthermore, this paper preliminarily explores how the identified puncture points are converted into coordinates for the robot’s end effector.