Computer Aided Medical Procedures, Johns Hopkins University, Baltimore, MD, USA.
Center for Micro-BioRobotics, Istituto Italiano di Tecnologia, Pontedera, Italy.
Int J Comput Assist Radiol Surg. 2017 Jun;12(6):1003-1011. doi: 10.1007/s11548-017-1561-1. Epub 2017 Mar 20.
We present the evaluation of the reproducibility of measurements performed using robotic ultrasound imaging in comparison with expert-operated sonography. Robotic imaging for interventional procedures may be a valuable contribution, but requires reproducibility for its acceptance in clinical routine. We study this by comparing repeated measurements based on robotic and expert-operated ultrasound imaging.
Robotic ultrasound acquisition is performed in three steps under user guidance: First, the patient is observed using a 3D camera on the robot end effector, and the user selects the region of interest. This allows for automatic planning of the robot trajectory. Next, the robot executes a sweeping motion following the planned trajectory, during which the ultrasound images and tracking data are recorded. As the robot is compliant, deviations from the path are possible, for instance due to patient motion. Finally, the ultrasound slices are compounded to create a volume. Repeated acquisitions can be performed automatically by comparing the previous and current patient surface.
After repeated image acquisitions, the measurements based on acquisitions performed by the robotic system and expert are compared. Within our case series, the expert measured the anterior-posterior, longitudinal, transversal lengths of both of the left and right thyroid lobes on each of the 4 healthy volunteers 3 times, providing 72 measurements. Subsequently, the same procedure was performed using the robotic system resulting in a cumulative total of 144 clinically relevant measurements. Our results clearly indicated that robotic ultrasound enables more repeatable measurements.
A robotic ultrasound platform leads to more reproducible data, which is of crucial importance for planning and executing interventions.
我们展示了使用机器人超声成像进行的测量的可重复性评估，与专家操作的超声检查进行比较。机器人成像在介入性手术中可能是一个很有价值的贡献，但需要可重复性才能在临床常规中被接受。我们通过比较基于机器人和专家操作的超声成像的重复测量来研究这一点。
在用户指导下，机器人超声采集分三个步骤进行：首先，机器人末端执行器上的 3D 相机观察患者，用户选择感兴趣的区域。这允许自动规划机器人轨迹。接下来，机器人按照规划的轨迹执行扫动运动，在此过程中记录超声图像和跟踪数据。由于机器人是顺应性的，因此可能会偏离路径，例如由于患者运动。最后，将超声切片组合以创建体积。通过比较之前和当前的患者表面，可以自动执行重复采集。
在重复图像采集后，比较基于机器人系统和专家采集的测量值。在我们的病例系列中，专家对 4 名健康志愿者的左右甲状腺叶的前后、纵向、横向长度进行了 3 次测量，每次测量提供了 72 个测量值。随后，使用机器人系统执行了相同的程序，总共获得了 144 个临床相关的测量值。我们的结果清楚地表明，机器人超声能够进行更可重复的测量。
机器人超声平台可实现更具可重复性的数据，这对规划和执行介入至关重要。
