Date of Award

1-18-2016

Degree Type

Thesis

Degree Name

M.Sc.

Department

Computer Science

First Advisor

Kent, Robert

Keywords

clinical data, data mapping, interoperability, LOINC, medical standards

Rights

CC BY-NC-ND 4.0

Abstract

The diversity in representation of medical data prevents straightforward data mapping, standardization and interoperability between the heterogeneous systems. We identify a specific problem, namely the need to achieve interoperability by applying a standard based data modeling approach to achieve a common platform that serves to improve the health data mapping of unstructured data and addresses ambiguity issues when dealing with health data from heterogeneous systems. In this thesis, we proposed an original Hybrid algorithm that identifies the attributes of data in heterogeneous systems based on critical medical standards and protocols and then performs semantic integration to form a uniform interoperable system. Also, efficient data modeling techniques are introduced for improving data storage and extraction. We tested the proposed algorithm with multiple data sets and compared the proposed approach with traditional data modeling approaches. We found that the proposed approach demonstrated performance improvements and reduction in data losses.

Share

COinS