Health Quest: A generalized clinical decision support system with multi-label classification

• Main Authors: Shahzeb Khan, Jawwad Ahmed Shamsi
• Format: Article
• Language: English
• Published: Elsevier 2021-01-01
• Series: Journal of King Saud University: Computer and Information Sciences
• Subjects: Clinical decision support system; Electronic Health Record; Deep learning; Natural language processing; GPUs; Big data
• Online Access: http://www.sciencedirect.com/science/article/pii/S1319157818306165

We propose Health Quest – a Deep Neural Network based Clinical Decision Support (CDS) system, which helps in diagnosing diseases through a patient’s Electronic Health Record (EHR). Health Quest is unique and distinct from existing CDS systems as it is a generalized system, which is not targeted to diagnosing a specific disease and can detect multiple diseases at a time. It utilizes a patient’s medical and family history and incorporates natural language processing to convert this information in a structured text for further processing. Multi-label classification is used to identify multiple diseases at a time. For improved accuracy and precision, Health Quest has been implemented on GPUs. We evaluate Health Quest on a dataset containing 5000 EHR records, with information of patients suffering from multiple diseases. We present the design and implementation of Health Quest and describe evaluation results. It is anticipated that Health Quest can be highly beneficial in prevention of mis-diagnosis of diseases – a major problem in medical sciences.