Department of Radiology, University of British Columbia, Vancouver, BC, V5Z 1M9, Canada.
Interventional Radiology Services, Vancouver General Hospital, G782, Jim Pattison Pavilion South, 899 West 12th Avenue, Vancouver, BC, V5Z 1M9, Canada.
Cardiovasc Intervent Radiol. 2019 Aug;42(8):1192-1198. doi: 10.1007/s00270-019-02207-8. Epub 2019 May 1.
To examine the feasibility of a novel technology platform that enables real-time touchless interaction with radiology images in both a simulated and an actual clinical setting.
This platform offers three different modes for image interaction. The gesture recognition mode uses a depth camera to detect the user's hand gestures which are translated to image manipulation commands. The light projection mode uses the same camera to detect finger point-and-tap movements above the icons which are projected on a surface to activate the commands. The capacitive sensing mode is enabled by a handheld, portable device, over which finger movements are detected by capacitive sensors to control the image review. Following initial feedback, light projection and capacitive sensing modes were selected for further testing by comparing with the conventional mode of image interaction in time trials for performing a series of standardized image manipulation tasks. Finally, the usability of the technology platform was examined in actual clinical procedures.
The light projection and the capacitive sensing modes were evaluated in the time trials and exhibited 60% and 71% reduction in time, respectively, relative to the control mode (p < 0.001). Clinical feasibility for this platform was demonstrated in three actual interventional radiology cases.
Accessing, navigating, and extracting relevant information from patient images intraprocedurally are cumbersome and time-consuming tasks that affect safety, efficiency, and decision-making during image-guided procedures. This study demonstrated that the novel technology addressed this issue by allowing touchless interaction with these images in the sterile field.
研究一种新的技术平台的可行性，该平台可在模拟和实际临床环境中实现与放射学图像的实时非接触式交互。
该平台提供了三种不同的图像交互模式。手势识别模式使用深度摄像机检测用户的手部手势，然后将其转换为图像操作命令。光投影模式使用同一摄像机检测手指在图标上的点按动作，以激活命令。电容感应模式由手持便携式设备启用，通过电容传感器检测手指在其上移动，以控制图像查看。在收到初步反馈后，选择光投影和电容感应模式进行进一步测试，方法是通过时间试验比较与传统的图像交互模式，以执行一系列标准化的图像操作任务。最后，在实际临床程序中检查了该技术平台的可用性。
在时间试验中评估了光投影和电容感应模式，与对照模式相比，它们分别减少了 60%和 71%的时间（p<0.001）。该平台在三个实际介入放射学病例中表现出了临床可行性。
在图像引导手术期间，术中访问、导航和提取患者图像中的相关信息是繁琐且耗时的任务，会影响安全性、效率和决策。本研究表明，该新技术通过在无菌区域实现对这些图像的非接触式交互解决了这个问题。
