PubMed HealthSearch

Biomedical subjects

H Pincé

Publications and source records attributed to H Pincé.

3 recordsLinked to original sources

Interpretative reporting and alarming based on laboratory data.

The utilisation of laboratory services for patient diagnosis and management involves many steps with both clinical and laboratory components. The clinical components include the decision to order a test, interpretation of the test results and actions taken on the basis of the results. The laboratory components on the other hand include receipt of the request, specimen collection, preparation and analysis, result entry, test result validation and verification and reporting of the results. In this paper, which is part of the OpenLabs project, we concentrate on the post-analytical applications which include interpretation and reporting of the laboratory results to the users in primary care and in high dependency care units. The final objective of the work described is to develop generic modules which can be integrated both with an Open laboratory information system architecture and existing laboratory information processing environment.

Acid-Base Imbalance

Computer aided interpretation of acid-base disorders.

This paper describes an expert system for the interpretation of acid-base disorders. The target users are residents in training in internal medicine, anaesthesia and intensive care medicine. The program is written in PROLOG and runs on a SUN 3/160 minicomputer. Evaluation of a learning set (N = 202) and a test set (N = 194) has proved that the system's accuracy is acceptable. As a result, the program has recently been put in routine clinical practice.

Acid-Base Imbalance

Computer aided phenotyping of dyslipoproteinemia.

An expert system was developed in order to obtain uniform and fast phenotyping and reporting of dyslipoproteinemia. PROLOG is used as knowledge representation formalism. The program is in daily use on an IBM PC AT, in the Laboratory of Clinical chemistry of the University Hospitals in Leuven. Evaluation has proved the system to be reliable and useful for the interpretation of lipoprotein disorders. Indeed accuracy figures of 98.4% and 95.2% were obtained in the learning (N = 315) and test set (N = 126) respectively.

Diagnosis, Computer-Assisted