The role of recommender system of tags in clinical decision support

Abstract

The widespread use of the Electronic Health Records EHRs has increasingly emerged in the healthcare industry. The structured and unstructured forms of EHRs are implemented in a Clinical Decision Support System CDSS. The CDSS is an health information technology system designed to provide healthcare professionals with clinical decision support. In this article, we aim to enhance the computer-aided diagnosis in medical imaging by recommending diseases for each patient’s medical image. We propose a recommender system of tags based on the tags co-occurrence, the graph of tags and the graph of the community of patients. The proposed approach is called MedicalRecomTags. The tags are the commonly used diseases or pathologies terms. The graphs, namely, the graph of tags and the graph of the community of patients, are derived by analyzing the annotated medical images. The experimental results show the effectiveness of the tag recommendation approach. In future works, the suggested tags will be evaluated by healthcare providers to affirm their relevancy. The intended online evaluation will enrich and enhance the recommender system of tags.