Hierarchical Distribution Analysis for Computer-assisted Classifications of Patient Records
- 1 December 1991
- journal article
- research article
- Published by SAGE Publications in Medical Decision Making
- Vol. 11 (4_suppl) , S94-S98
- https://doi.org/10.1177/0272989x9101104s18
Abstract
The many existing hierarchical classifications present opportunities for computer-aided classification of medical data. The authors evaluate lexical term mapping followed by hierarchical node activation to determine the semantic locality of diagnoses in the American College of Radiology (ACR) classification. The activated hierarchy exposes conceptual structures behind lexically matched categories and markedly narrows the alternatives for human review. Several distribution analyses are evaluated for narrowing the candidate search space, to reduce ambiguity. The methods were tested on a prototype system, using a set of 78 samples of Mayo radiology diagnoses; their relative effectiveness is illustrated.Keywords
This publication has 3 references indexed in Scilit:
- SAPHIRE—An information retrieval system featuring concept matching, automatic indexing, probabilistic retrieval, and hierarchical relationshipsComputers and Biomedical Research, 1990
- A vocabulary for medical informaticsComputers and Biomedical Research, 1987
- Automated medical office recordsJournal of Medical Systems, 1987