Developed an AI model that predicts potential serious case of COVID-19Author : 관리자Date : 2021.04.21Hit : 151
Developed AI technology for early identification of potential serious case of COVID-19
- Professor Jinseok Lee's research team at CUK Department of Artificial Intelligence developed an AI model that predicts potential serious case of COVID-19
- Predicts a patient's severity immediately upon entering their clinical data at an early stage Published a website incorporating the AI technology
- Factors that can deteriorate severity: old age, low lymphocyte level, low platelet count, shortness of breath at an early stage
Professor Jinseok Lee's research team at the Catholic University of Korea Department of Artificial Intelligence developed an artificial intelligence (AI) model capable of early prediction of the possibility of serious case of a patient with coronavirus disease 2019 (COVID-19). A website incorporating this artificial intelligence model can identify potential serious cases immediately upon entering the clinical data of patients at an early stage.
This research was conducted based on the clinical data of 5,628 COVID-19 patients, collected from around 100 medical institutions in Korea through the Korea Disease Control and Prevention Agency. According to the research analysis by professor Jinseok Lee's research team, factors such as old age, low lymphocyte level, low platelet count and shortness of breath at an early stage can raise the likeliness of severity, irrelevant of early symptoms.
Based on the findings from the clinical results, the team sequenced the severity factors and developed a Deep Neural Network (DNN) artificial intelligence model. They have perfected a highly reliable artificial intelligence model with prediction accuracy as high as 90.4% by collaborating with 3 medical institutions: Jeju Emergency Center, ASAN Medical Center, and Wonkwang University School of Medicine.
To make this artificial intelligence model available for all medical teams and researchers worldwide, professor Jinseok Lee and his research team published a dedicated website (http://kcovidnet.site). You can check a patient's likeliness of developing serious case of COVID-19 by simply entering their clinical data on the website.
Professor Jinseok Lee of the Catholic University of Korea Department of Artificial Intelligence said, “Predicting the severity of COVID-19 that can possibly lead to serious case and providing intensive treatment as a preemptive measure irrelevant of the early symptoms can help reduce mortality as well as efficiently distribute medical resources”, “We hope this artificial intelligence model will assist medical teams and researchers worldwide to easily identify COVID-19 patients and take preemptive measures.”
Meanwhile, this research outcome by professor Jinseok Lee and his team at CUK Department of Artificial Intelligence has been published in the April issue of the Journal of Medical Internet Research (JMIR), an eminent academic journal in the top 5% of the medical information field.