Electronic Health Record Mining for Early Disease Risk Prediction

Dhanaraj Sathiri

Abstract: Mining electronic health records (EHR) for patterns predictive of the onset of diseases and clinical events has the potential to generate knowledge that supports the identification of at-risk patients in the early stages of disease evolution. The HPC-ML-HAZARD project aims to design, build, and evaluate predictive models that are applicable in clinical settings. The early identification of patients at risk facilitates interventions, system-level management, and ultimately the prevention of excess morbidity and mortality. Furthermore, early risk prediction has potential implications not limited to the patients identified, including informing and optimizing the allocation of hospital resources. The use of EHRs for early prediction of diseases has received relatively little attention compared to other uses, especially considering the potential clinical significance. The prospect of identifying patients at risk before the onset of disease or clinical complications is enticing, as it provides an opportunity to implement preventive measures, support systems, and clinical research to clearly identify preclinical disease stages.

Keywords: Electronic health records (EHRs), Early warning systems, Risk stratification, Impact on clinical outcomes, Risk evolution

How to Cite?: Dhanaraj Sathiri, "Electronic Health Record Mining for Early Disease Risk Prediction", Volume 9 Issue 12, December 2020, International Journal of Science and Research (IJSR), Pages: 1979-1987, https://www.ijsr.net/getabstract.php?paperid=MS2012181310, DOI: https://dx.doi.org/10.21275/MS2012181310