Interface of a commercial electrocardiogram system to HELP;
The Electrocardiogram (ECG) Department at the Latter-Day-Saints Hospital has used a computerized ECG system for more than 15 years. In this automated system, all the ECG data were analyzed by a set of Health Evaluation Logic Processing (HELP) frames and the resulting interpretations were stored in the patient data base. In 1987, the Department replaced this system with the Marquette Universal System for Electrocardiography (MUSE). As a stand alone system, MUSE stores all its ECG interpretations in its own data base. Since the HELP system serves as the information center for the Hospital, it is necessary to establish an interface between the HELP and the MUSE systems so that the MUSE ECG interpretations can be stored in HELP and become available to the clinical personnel. To integrate the MUSE system to HELP, one first faces the challenge of terminology difference between these two systems. According to the degree of compatibility among the terms used, there exist three categories of the MUSE interpretations. Different strategies were used in defining the Pointer to TeXT (PTXT) codes for these three categories of MUSE interpretations. In the process of constituting the PTXT representations for the MUSE system, care was taken to avoid duplicating existing codes in the HELP data dictionary. The second issue in interfacing the MUSE system to HELP lies in understanding the MUSE statements. A MUSE statement may contain different interpretations. Therefore, if a MUSE statement is to be stored in the HELP system, the interpretations constituting this statement must be understood so that their corresponding PTXT codes can be stored. In order to do this, a parsing algorithm was designed to detect different interpretations used in a statement and store their PTXT representations to HELP. After the implementation of the interface software, it was found that the software was constantly ready to capture the MUSE data into the HELP system. In addition, all the MUSE EGG statements, after being processed by the parsing algorithm, had been transferred to semantically corresponding interpretations. These interpretations were stored as patient records and were available to the reviewing physicians throughout LDS Hospital.
University of Utah
Artificial Intelligence; Electrocardiography; Medical Informatics Computing; Monitoring, Physiologic; Signal Processing, Computer-Assisted; Software;
University of Utah
Relation-Is Version Of
Digital reproduction of “Interface of a commercial electrocardiogram system to HELP Spencer S. Eccles Health Sciences Library.