An intelligent decision support system: prediction of hospitalization and length of stay, and similar medical records retrieval for proper patients arrangement in an emergency department.
Summary The core concept of the decision support system is to establish machine learning-based predictive models and enable their interpretability to support physicians making clinical decisions in practice. The developed technique is part of the capstone project named "Smart Emergency Department," sponsored by the Ministry of Science and Technology. The system comprises NTUH (National Taiwan University Hospital) EMR (Electronic Medical Record) importance analysis, accurate clinical quality indicator prediction, and similar medical record retrieval. It is expected to improve the patients' flow and alleviate emergency department crowding. It has been verified in retrospective studies and will officially enter the clinical trial phase this year.
Over one million EMRs records and plenty of features in the clinical timeline for each sample have been extracted through the retrospective research. Collected EMRs were distributed across three regional hospitals. The quantity and quality of the current database are remarkable and convincible for paper publication. The best performance of the model achieves 0.94 of AUC in predicting both hospitalization and length of stay, which is comparable to the state-of-the-art in the literature. The newly proposed technique, unsupervised quantification of similar medical records and retrieval, is an innovative approach for the tree-based model and also provides interpretability. Outcomes of subsequent clinical trials have the potential to publish results in high-impact clinical journals.
The application domain of the technique is expected in the field of providing medical services. After complete prospective study and obtain clinical evidence successfully, the experience and parameters accumulated during the development period will serve for the promotion of smart healthcare and the foundation of commercial design, which can help system manufacturers to formulate compatible standards for the operation of intelligent modules and assist in the communication of medical records. In addition, help set up computing service equipment locally and privately. After smart healthcare becomes more popular, small equipment factories can also develop peripheral products for smart healthcare, forming a supply chain of related industries.
Technical Film
Keyword Smart Emergence Department#Electronic Medical Record Machine Learning Clinical index prediction Patient Similarity Similar EHR Retrieval
Research Project
Research Team
More like this
Provide the latest information of AI research centers and applied industries