Ruchelman M, Krishnamurthy K, Hostetler W, Peterson L L, Jungmann J, Doller H J
VA Medical Center, Dallas, Texas.
Proc Annu Symp Comput Appl Med Care. 1992:816-7.
While medical expert systems helped demonstrate that artificial intelligence was possible, few medical systems have been heralded as practical successes. We believe that expert systems will be practical successes if they cost effectively handle most of a physician's workload (i.e., routine care). To accomplish this goal, technology must appear invisible to the user; the system must be intuitive and anticipate users' needs. "Epileptologists' Assistant" is an example of our approach of combining a graphical user interface with an expert system and data base in a system to help in a routine specialty clinic. The goal is for two nurses and a physician to handle the workload of three physicians while increasing the quality of care. The current system reduces physician time by 66%. Our ultimate goal is to create a unified family of systems for medical specialties.
虽然医学专家系统有助于证明人工智能是可行的,但几乎没有医学系统被视为取得了实际成功。我们认为,如果专家系统能够经济高效地处理医生的大部分工作量(即常规护理),那么它们将会取得实际成功。为实现这一目标,技术必须对用户来说是无形的;系统必须直观且能预见到用户的需求。“癫痫专家助手”就是我们将图形用户界面与专家系统及数据库相结合应用于系统,以协助常规专科诊所工作的一个实例。目标是让两名护士和一名医生完成三名医生的工作量,同时提高护理质量。当前系统将医生的时间减少了66%。我们的最终目标是为各个医学专科创建一个统一的系统家族。