Medical Motion Planning Dataset (Med-MPD)

Medical Motion Planning Dataset (Med-MPD) 是由美国国家卫生研究院支持创建的公开数据集，专注于评估医疗应用中的运动规划器。该数据集包含15个真实的临床场景，分布在肺、肝和脑三个器官中，用于评估微创医疗机器人的运动规划器。数据集的创建旨在为医疗机器人运动规划提供一个标准化的评估框架，推动相关研究的发展，并减轻生成医疗评估数据的负担。数据集内容包括临床相关的环境描述、临床动机和评估标准，适用于多种医疗程序，如活检、消融和药物输送。

Medical Motion Planning Dataset (Med-MPD) is a publicly available dataset developed with support from the U.S. National Institutes of Health, focusing on evaluating motion planners for medical applications. This dataset includes 15 real clinical scenarios across three organ systems: lung, liver, and brain, which are used to evaluate motion planners for minimally invasive medical robots. The dataset was developed to provide a standardized evaluation framework for medical robot motion planning, advance relevant research, and reduce the burden of generating medical evaluation data. It encompasses clinically relevant environment descriptions, clinical rationales, and evaluation criteria, applicable to a wide range of medical procedures such as biopsy, ablation, and drug delivery.

搜集汇总

背景与挑战

背景概述

Medical Motion Planning Dataset (Med-MPD)是一个用于评估运动规划器和微创机器人设计的临床数据集，包含来自患者解剖结构的肺部、肝脏和大脑环境，每个环境提供分割文件和临床任务坐标。该数据集基于公开医学影像数据（如TCIA），以.nii.gz格式提供，适用于医疗机器人运动规划研究，并已扩展为Med-RAD项目。