Controlling Anesthesia Hardware With Simple Hand Gestures: Thumbs Up or Thumbs Down?
Controlling Anesthesia Hardware With Simple Hand Gestures: Thumbs Up or Thumbs Down?
复制标题
用简单的手势控制麻醉硬件：竖起大拇指还是竖起大拇指？
DOI：
10.1213/ane.0000000000005071
复制
发表时间：
2021-07-01
影响因子：
5.7
通讯作者：
Connor CW
中科院分区：
文献类型：
--
作者：
Owens GE;Connor CW
现代消费电子设备和汽车通常由能感知身体手势和语音指令的界面控制。相比之下，病人监护仪和麻醉设备通常配备有安装在面板上的按钮、拨号盘和键盘。麻醉中越来越多地使用非接触式手势界面，可以通过更直观、更迅速地控制设备，以及减少表面污染率，提高患者的安全性。设计了一种新型的基于手势的控制器，并将其改造为标准的GE Solar 8000M患者监护仪。由于商业化生产的临床设备上的专有输入控制系统被严格控制，这种类型的技术创新是罕见的。尽管如此，我们假设麻醉师会发现非接触式手势界面使用起来很简单。利用毫米波雷达传感器，开发了一种基于手势的界面系统来控制Solar 8000M患者监护仪。该系统被编程为检测非接触式“旋转”和“按压”手势，通过实现虚拟调节旋钮进行界面控制来控制患者监护仪。50名麻醉师测试了一个原型界面，并通过填写一份包含修改李克特量表的简短问卷来评估可用性。这些评估是在非患者护理环境中进行的，因此受访者在评估期间不会有不利的任务负荷，也减轻了使用这种新型患者管理界面的任何伦理或安全问题。麻醉硬件通过手势传感器上方的两个不同手势进行可靠的控制。基于手势的界面普遍受到麻醉医师的欢迎(8.09（置信区间8.06,8.12），他们更喜欢简单的“按压”手势，而不是“旋转”手势(8.45 （8.39,8.51)vs 7.73(7.67,7.79)， 10分制，p = 0.005）。每位麻醉师对两种手势的偏好评分之间存在很强的相关性（Pearson r = 0.49 (0.25,0.68), p < 0.001）。高级培训水平（住院医师、研究员、1-10年主治医师、10年主治医师）与两种手势的偏好得分无关(“按压”的Spearman ρ = - 0.02 （- 0.30,0.26), p = 0.87；“旋转”的Spearman ρ = 0.08 (- 0.20,0.35), p = 0.58）使用手势感应来控制麻醉设备受到一群麻醉医师的好评。尽管更简单的“按”手势比“旋转”手势更受欢迎，但受访者内部的相关性表明，整体上对手势的偏好效果更强。可接受性与麻醉经验水平无负相关。手势感应是一个很有前途的新领域，可以简化和改善麻醉师与麻醉工作站之间的交互。
Modern consumer electronic devices and automobiles are often controlled by interfaces that sense physical gestures and spoken commands. In contrast, patient monitors and anesthesia devices are typically equipped with panel-mounted buttons, dials and keyboards. The increased use of non-contact gesture-based interfaces in anesthesia may improve patient safety through more intuitive and prompter control of equipment, and also through reduced rates of surface contamination. A novel gesture-based controller was designed and retrofitted to a standard GE Solar 8000M patient monitor. This type of technical innovation is rare, due to closely-held proprietary input control systems on commercially produced clinical equipment. Nevertheless, we hypothesized that anesthesiologists would find a contactless gesture interface straightforward to use. A gesture-based interface system was developed to control a Solar 8000M patient monitor using a millimeter wave radar sensor. The system was programmed to detect non-contact “rotate” and “press” gestures to control the patient monitor by implementing a virtual trim knob for interface control. Fifty anesthesiologists tested a prototype interface and evaluated usability by completing a short questionnaire incorporating modified Likert scales. These evaluations were performed in a non-patient-care environment so that respondents were not adversely task-loaded during assessment, also allaying any ethical or safety concerns regarding use of this novel interface for patient management. Anesthesia hardware was controlled reliably with two distinct gestures above the gesture sensor. The gesture-based interface generally was well received by anesthesiologists (8.09 (Confidence Interval 8.06,8.12) on a 10-point scale), who preferred the simpler “press” gesture to the “rotate” gesture (8.45 (8.39,8.51) vs 7.73 (7.67,7.79) on a 10-point scale, p = 0.005). The correlation between the preference scores for the two gestures from each anesthesiologist was strong (Pearson r = 0.49 (0.25,0.68), p < 0.001). Advancing level of training (Resident, Fellow, Attending 1-10 years, Attending > 10 years) was not correlated with preference scores for either gesture (Spearman ρ = −0.02 (−0.30,0.26), p = 0.87 for “press” and Spearman ρ = 0.08 (−0.20,0.35), p = 0.58 for “rotate”). The use of gesture sensing for controlling anesthesia equipment was well-received by a cohort of anesthesiologists. Even though the simpler “press” gesture was preferred over the “rotate” gesture, the intra-respondent correlation indicates that the preference for gestures as a whole is the stronger effect. No adverse relationship was found between acceptability and anesthesia experience level. Gesture sensing is a promising new area to simplify and improve the interaction between the anesthesiologist and the anesthesia workstation.