Research team led by Professor Kyung-Soo Jeong, Department of Respiratory Medicine, Severance Hospital of Yonsei University
Posted 2024.01.16 19:10 Views 3 Posted 2024.01.16 19:10 Modified 2024.01.16 15:34 Views 3
An artificial intelligence (AI) model has been developed that can accurately and quickly diagnose sepsis and even predict prognosis. [사진=클립아트코리아]An artificial intelligence (AI) model has been developed that can accurately and quickly diagnose sepsis and even predict prognosis. Sepsis is a disease that causes damage to major organs due to the body’s abnormal response to microbial infections. The mortality rate due to severe sepsis increases to 35% and up to 60% if accompanied by septic shock.
The research team of Professor Yu-Rang Park from the Department of Biomedical Systems Information at Yonsei University College of Medicine and Professor Gyeong-Soo Jeong from the Department of Respiratory Medicine at Severance Hospital has developed an AI model that can diagnose sepsis and predict prognosis using the developed 3D image data of CD8 T cells that kill virus-infected cells or tumor cells. Accuracy is 99%.
Because the immune response to sepsis is complex and varies from patient to patient, early diagnosis and timely action are important. Because it affects multiple organs rapidly, the chance of death increases if treatment is delayed.
Representative biomarkers currently used to diagnose sepsis, such as C-reactive protein (CRP) and procalcitonin (PCT), have delayed diagnosis, resulting in delayed responses. Furthermore, biomarkers such as interleukin-6 (IL-6), an inflammatory marker, lacked standardization, making the interpretation of diagnostic results difficult. Therefore, it was necessary to discover new biomarkers.
The research team examined whether the diagnosis and prognosis of sepsis could be predicted using CD8 T cell image data and artificial intelligence models. CD8 T cells were isolated from blood samples of eight people in the sepsis recovery group (including those who died), and images were taken. Filming was conducted at three time points: when septic shock was diagnosed, when septic shock resolved, and before discharge, and a holotomographic microscope was used. Holotomography can quickly and reliably obtain 3D images of living immune cells without a staining process that affects changes in cell structure.
Images taken at each time point were compared and analyzed with images from 20 healthy control subjects using an AI classification model. At this point, the images obtained during the diagnosis of septic shock were used to evaluate the possibility of diagnosing septic shock, and the images obtained during the diagnosis of septic shock in the groups of surviving and deceased patients were used to predict the prognosis of the septic shock.
The prediction performance of the AI model was analyzed using the receiver operating characteristic curve (AUROC) index. AUROC means “area under the ROC curve”. It is a statistical technique that indicates the diagnostic accuracy of a specific test tool and is used as a performance evaluation indicator for artificial intelligence models. Typically, the closer the area is to 1, the better the performance, and if the area is 0.8 or higher, it is rated as a high-performance model.
As a result of the analysis, when only one CD8 T cell image was used to diagnose sepsis, the prediction accuracy of the AI model (AUROC) was 0.96 (96%), and when two CD8 T cell images were used , performance was above 0.99 (99%) It seemed. The prognosis prediction model also showed an accuracy of 0.98 (98%) using a single-cell image and showed high performance above 0.99 (99%) when using two-cell images.
Professor Gyeongsoo Jeong, who led the research, said: “Through this study, we were able to identify the role of 3D images of CD8 T cells as a biomarker for sepsis.” sepsis patients through an artificial intelligence model, “We expect it to be able to help make appropriate treatment decisions for each patient.”
The results of this study were published in the latest issue of the international academic journal Light: Science&Application.
Journalist Lim Jong-eon
“Copyright ⓒ ‘Honest Knowledge for Health’ Comedy.com ( / Unauthorized reproduction and redistribution prohibited”
2024-01-16 10:11:23
#Sepsis #diagnosis #save #lives #quickly #possible.. #secret #increasing #accuracy