JMIR Publications

ABSTRACT

Background:

According to the World Health Organization, mental disorders currently represent a global public health priority. However, in this area, health managers are forced to interpret complex data obtained from dispersed sources, with low integration, varied accuracy, and often accessible only by specific information systems, resulting in reduced data quality and integrity. This scenario generally prevents the extraction of knowledge and its use as a reference to offer decision-making support for health professionals and managers in clinical, operational, and administrative processes. Then, interoperability between health information systems is essential to establish the ability to communicate, exchange and reuse shared information. The Semantic Web is a validated fundamental structure to ensure semantic interoperability and the integration of dispersed and isolated datasets through computational biomedical ontologies.

Objective:

Given the importance of mental health and the need for quality data in this area, this study aims to specify and develop a knowledge-based platform through Semantic Web technologies and standards to integrate, consume, analyze, and visualize mental health information in a public health care network.

Methods:

It is applied research. Therefore, we carried out a literature review on the mental health domain and the Semantic Web area. From these reviews, we acquired and structured the knowledge needed to build a computational biomedical ontology for the mental health area and an infrastructure for its use, with three main components: Resource Description Framework, Ontology Web Language, and the SPARQL Protocol and RDF Query Language. Then, we used public data from the Psychosocial Care Network of Ribeirão Preto, in line with the guidelines of the Ministry of Health of Brazil, the World Health Organization, and the Diagnostic and Statistical Manual of Mental Disorders, to populate and test the created semantic platform. Then, we set up an environment to perform analysis on this dataset.

Results:

As a result of this study, we structured the terminology and standard vocabulary in mental health according to documents from official health agencies offering a structure to promote semantic interoperability in mental health. We developed and published a biomedical ontology named Mental Health Management Ontology on the web and a proposal for a computational architecture for its use. Through the information stored in this application, we build a solid and consistent knowledge base through which we perform and make available analyzes and extracts of clinical and managerial knowledge.

Conclusions:

The designed structure is a technical contribution that aims to minimize the complexity of mental health data through highly expressive relationships, standardization of concepts and vocabularies, interoperable technologies, and consistent logical rules. To promote improvements in evidence-based decision-making, improve the service provided to the population, and optimize resource use, benefiting patients, health professionals, and managers.