Notice

List

Smartphone-based AI for High-Sensitivity Virus Diagnosis in the Field

  • admin
  • 2023-05-15
  • 3277
-Professor Jeonghoon Lee's team develops a new technology using smartphone deep learning to diagnose and predict virus diseases in the field-

-Developed a proprietary deep learning algorithm-based high-sensitivity on-site diagnostic technology using smartphones-

-Blind evaluation results for COVID-19 show a significant increase in sensitivity from 72% for the general public to 100%-

-The sensitivity for difficult-to-diagnose infections in the early stages significantly increased from 51% for the general public to 91% when AI was applied-

-Ensured versatility in diagnosis using smartphones or rapid kits that are widely available on the market-

-Presented a new paradigm for on-site diagnosis that can judge and suggest the early diagnosis of new and variant viruses, the progress of diseases, and the need for isolation-

-Published in Nature Communications (IF: 17.69)-

 

Professor Lee Jung-hoon's research team (Department of Electrical Engineering) has successfully developed an AI diagnostic technology based on smartphones through joint research with Kellss Co., Ltd., Professor Lee Ki-baek's team at Kwangwoon University, Professor Jo Sung-yeon's team at Seoul St. Mary's Hospital, Professor Yoo Yong-kyung's team at Catholic Kwandong University, and Professor Yoon Dae-seong's team at Korea University. Through this technology, it is possible to perform highly sensitive on-site diagnosis using only a smartphone and a rapid kit without any external devices.

 

This technology has been transferred to Kels Co., Ltd., and commercialization is planned through approval and certification processes such as app/algorithm optimization, and the U.S. Food and Drug Administration (FDA) and the Korean Food and Drug Administration (KFDA). This research was supported by the Bio-Medical Technology Development Program of the National Research Foundation of Korea (No. 2021M3E5E3080743) and published in Nature Communications, a top-tier journal published by Nature Portfolio (IF: 17.69).

 

Web link: https://www.nature.com/articles/s41467-023-38104-5

 

<왼쪽부터 연구책임자 이정훈 교수(광운대), 이승민(박사과정), 김선목(박사과정), 윤대성 교수(고려대), 이기백 교수(광운대), 유용경 교수(가톨릭관동대)>？

<From left to right, the research team consists of Professor Jeonghoon Lee (Kwangwoon University), Seungmin Lee (PhD candidate), Sunmok Kim (PhD candidate), Professor Daesung Yoon (Korea University), Professor Kiback Lee (Kwangwoon University), and Professor Yongkyung Yoo (Catholic Kwandong University)>

스마트폰 AI 기술 (SMARTAI-LFA)을 활용한 진단기술 시스템과 알고리즘. 본 시스템을 적용함으로서, 단순히 키트를 스마트폰으로 촬영하는 것만으로도 초감도 현장진단이 가능함을 확인하였음. 블라인드 평가결과 민감도가 72% (일반인) 에서 100%로 크게 증가 하였으며, 진단 어려운 감염 초기의 민감도가 51% (일반인) 에서 91%로 크게 증가하는 것을 확인하였음. 시중 판매되는 스마트폰과 래피드 키트와 호환 및 진단이 가능함을 확인하였음. 향후 신변종 감염병의 조기 진단의 가능성과, 병의 진행 및 격리 등의 AI 기술 적용 가능성을 제시하였음.？ 

Image description: Diagnostic technology system and algorithm using smartphone AI technology (SMARTAI-LFA). By applying this system, it was confirmed that ultra-sensitive on-site diagnosis is possible by simply photographing the kit with a smartphone. The blind evaluation results showed that the sensitivity increased significantly from 72% (general) to 100%, and the sensitivity in the early stages of infection, which is difficult to diagnose, increased significantly from 51% (general) to 91%. It was confirmed that it is compatible and can be diagnosed with commercially available smartphones and rapid kits. The possibility of early diagnosis of new variants of infectious diseases and the application of AI technology for the progression of the disease and isolation were presented.
List