SBIR Phase I: Robotic Scrub Technician

这个小企业创新研究（SBIR）第一阶段项目建议为手术室的机器人擦洗技术员开发认知结构的初始组件。擦洗技术员将保持一个器械托盘，当外科医生要求时将器械交给外科医生，当外科医生用完器械时取出器械。机会在于降低医院人员成本，缓解严重的护理人员短缺问题。问题是要让机器人具备与经验丰富的人工手术技师相当的态势感知能力，尤其是在预测外科医生下一个要求的能力上。该项目的目标是确定使用人工智能和统计技术来提供这种能力的可行性。研究方法将比较时间序列预测和分类的选定技术，使用实际手术的数据集。候选技术将包括一阶到三阶马尔可夫方法、n序列匹配、神经网络和基于模糊集的推理。预计最终产品将是一个预测引擎，使临床版本的机器人表现得和经验丰富的人类清洁技术人员一样好，甚至更好。该项目的商业应用是在生物医学设备和仪器领域。

This Small Business Innovation Research (SBIR) Phase I project proposes to develop the initial component of a cognitive structure for a robotic scrub technician in the operating room. The scrub technician will maintain a tray of instruments, handing an instrument to the surgeon when requested and retrieving the instrument when the surgeon is finished with it. The opportunity is to reduce hospital personnel costs and alleviate a critical nursing shortage problem. The problem is to provide the robot with situational awareness comparable to an experienced human scrub technician, particularly in the ability to anticipate the surgeon's next request. The objective of the project is to determine the feasibility of using artificial intelligence and statistical techniques to provide this ability. The research method will compare selected techniques for time series prediction and classification, using data sets from actual surgeries. The candidate techniques will include first to third order Markov methods, N-sequence matching, neural networks and fuzzy set based inference. The end product is expected to be a prediction engine that will enable the clinical version of the robot to perform as well or better than an experienced human scrub technician. The commercial application of this project is in the area of biomedical devices and instrumentation.