Motion detection technology as a tool for cardiopulmonary resuscitation (CPR) quality improvement.

The most popular method of training in basic life support nd AED remains instructor-led training courses. Recent reviews rovide good evidence to support alternative methods of trainng including lay instructors, self-directed learning (web, video, oster) and CPR feedback/prompt devices.1,2 We report here he results of the prototype testing of the previously described ini-VREM project.3 In brief we used the Kinect© sensor system to onitor the performance of CPR on a manikin. Kinect© is a motion ensing input device made by Microsoft for the Xbox 360 game onsole. It enables users to control and interact with the Xbox 60. It does this through a natural user interface which recognises estures and spoken commands without the need to touch a ame controller. Kinect’s camera is driven by both hardware and oftware. It does two things: it generates a three-dimensional mage of the objects in its field of view and recognises human eings among those objects. The camera transmits invisible ear-infrared light and works like sonar. At this point, both the inect’s hardware—its camera and IR-light projector—and its rmware called “middleware” are operating. The Kinect© has an n-board processor which uses algorithms to process the data o render the three-dimensional image. The middleware can istinguish human body parts, joints and movements, as well s distinguishing individual human faces from one another. The inect© sensor can detect objects within a working distance of .2–3.5 m. The horizontal field of the Kinect© sensor is 87 cm, and he vertical field is 63 cm, resulting in a resolution of just over .3 mm (0.051 in.) per pixel. With this technology, Kinect© can istinguish objects depth within 1 cm and their height and width ithin 3 mm. This resolution is sufficient for quality CPR analysis chest compression rate and depth). Recently, Microsoft released non-commercial Kinect© software development kit (SDK) for indows. It enables the academic and enthusiast communities asy access to the capabilities offered by the Microsoft Kinect© evice connected to computers running the Windows 7 operating ystem.4 Our preliminary tests demonstrate the feasibility of the inect© sensor to monitor CPR performance and can be viewed on -tube.5 Kinect© was able to reproducibly track limb movement uring external chest compression and demonstrated excellent ynchrony with simultaneous video recordings. We are now