Khitrov Maxim Y, Rutishauser Matthew, Montgomery Kevin, Reisner Andrew T, Reifman Jaques
Bioinformatics Cell, TATRC, USAMRMC, Frederick, MD 21702, USA.
Annu Int Conf IEEE Eng Med Biol Soc. 2009;2009:3417-20. doi: 10.1109/IEMBS.2009.5332403.
The unavailability of a flexible system for realtime testing of decision-support algorithms in a pre-hospital clinical setting has limited their use. In this study, we describe a plug-and-play platform for real-time testing of decision-support algorithms during the transport of trauma casualties en route to a hospital. The platform integrates a standard-of-care vital-signs monitor, which collects numeric and waveform physiologic time-series data, with a rugged ultramobile personal computer. The computer time-stamps and stores data received from the monitor, and performs analysis on the collected data in real-time. Prior to field deployment, we assessed the performance of each component of the platform by using an emulator to simulate a number of possible fault scenarios that could be encountered in the field. Initial testing with the emulator allowed us to identify and fix software inconsistencies and showed that the platform can support a quick development cycle for real-time decision-support algorithms.
在院前临床环境中，缺乏用于实时测试决策支持算法的灵活系统限制了它们的应用。在本研究中，我们描述了一种即插即用平台，用于在创伤伤员转运至医院途中实时测试决策支持算法。该平台将收集数字和波形生理时间序列数据的标准护理生命体征监测器与坚固的超便携个人计算机集成在一起。计算机对从监测器接收到的数据进行时间戳记和存储，并对收集到的数据进行实时分析。在实地部署之前，我们通过使用模拟器模拟一些在实地可能遇到的故障场景，评估了平台各组件的性能。使用模拟器进行的初步测试使我们能够识别并修复软件不一致问题，并表明该平台可以支持实时决策支持算法的快速开发周期。
