PubMed HealthSearch

Biomedical subjects

F Naeymi-Rad

Publications and source records attributed to F Naeymi-Rad.

4 recordsLinked to original sources

Informatics Workup.

We introduce the concept of a Medical Informatics Workup performed by fourth year medical students working in a busy inner-city Emergency Room. These students use portable computers (Macintosh PowerBook 170s connected to a removable cartridge hard drive and CD-ROM drive) to do the patient workups. The PowerBook 170 contains the automated medical record entry software (IMR-E), five expert system software packages, and a program that allows the PowerBook to emulate a PC-compatible computer. With this configuration the student has a portable system that allows for the creation of a computerized medical record at the patient's bedside, along with the ability to analyze the data and generate a list of differential diagnoses.

Angina Pectoris

A test selection module for MEDAS.

A test selection module to recommend proper tests to the physician during the diagnostic process has been added to MEDAS (the Medical Emergency Decision Assistance System). Test selection is an important part of the diagnostic process. This module can use either diagnostic hypotheses set up by the physician or derive its own diagnostic hypotheses from the differential diagnosis. This module recommends pertinent questions to be asked about the patient's history and physical examination and recommends laboratory and radiological tests. The recommended tests are presented in two different ways: as individual tests and as groups of sequential tests.

Algorithms

A feature dictionary supporting a multi-domain medical knowledge base.

Because different terminology is used by physicians of different specialties in different locations to refer to the same feature (signs, symptoms, test results), it is essential that our knowledge development tools provide a means to access a common pool of terms. This paper discusses the design of an online medical dictionary that provides a solution to this problem for developers of multi-domain knowledge bases for MEDAS (Medical Emergency Decision Assistance System). Our Feature Dictionary supports phrase equivalents for features, feature interactions, feature classifications, and translations to the binary features generated by the expert during knowledge creation. It is also used in the conversion of a domain knowledge to the database used by the MEDAS inference diagnostic sessions. The Feature Dictionary also provides capabilities for complex queries across multiple domains using the supported relations. The Feature Dictionary supports three methods for feature representation: (1) for binary features, (2) for continuous valued features, and (3) for derived features.

Artificial Intelligence