‘Trying to take the lead in artificial intelligence-based medical market’
A paper on artificial intelligence research is published
in internationally renowned academic journal
Mediplex Sejong Hospital began its move to take the lead in artificial intelligence-based medical market in cooperation with VUNO, a medical data analysis company based on artificial intelligence.
The paper on deep-learning-based triage and acuity score, which jointly developed by Mediplex Sejong Hospital and VUNO based on artificial intelligence was published in PLOS ONE, an internationally renowned medical journal (Kwon JM, Lee Y, Lee Y, et al. Validation of deep-learning-based triage and acuity score using a large national dataset, PLOS ONE, 2018 Oct.)
It is an artificial intelligence tool developed using a big data of 12 million patients who visited emergency room provided to Central Emergency Medical Center and accuracy of the tool was verified by multi-agency researches after development.
It predicts the risk of death, necessity of treatment for critical patients, and necessity of inpatient care with very high accuracy by using only initial information of emergency room patients and is 1.5 times more accurate than Korean Triage and Acuity Scale (KTAS). Existing tool consumes much time as manual calculation is required based on subjective criteria such as level of pain, which results in differences among medical team and cannot be applied during first aid state.
Deep-learning-based triage and acuity score, which is newly developed, can be applied to patients in transit as it provides prediction using only initial information such as age, sex, major symptoms, injury, initial vital signs and does not require manual work as calculation is available immediately after inputting information. It can be used automatically in emergency situation as it is linked with electronic medical records and can be applied to patients who are being transferred by ambulance in a convenient and accurate manner. In addition, it maintains accuracy stably as only objective information is used, enabling identification and prediction of patient condition with high accuracy.
The paper on prediction of cardiac arrest was published in Journal of the American Heart Association (JAHA) in last July and Medical Sejong Hospital has conducted researches in cooperation with other domestic hospitals. This time, Medical Sejong Hospital developed an artificial intelligence which can be utilized in emergency room, fire fighting and relief stages. We will lead the research and development for artificial intelligence technology and research by conducting follow-up researches to make sure that this tool will be used for patient treatment and patient safety.