Current Result Document :
ÇѱÛÁ¦¸ñ(Korean Title) |
LSTM À» »ç¿ëÇÑ ÀÔ¿ø ±â°£ Áß ¾à¹° À¯Àμº °£¼Õ»óÀÇ ¹ß»ý ¿¹Ãø |
¿µ¹®Á¦¸ñ(English Title) |
Predicting In-Hospital Drug-Induced Liver Injuries Using Long-Short Term Memory |
ÀúÀÚ(Author) |
µµ¸£ÁñįµÑ±º
¹éÁÖ¿ø
Á¶È¿Á¤
±è¼ø¼±
ÇÑÁöÀº
Á¤À翬
Dulguun Dorjlkham
Juwon Baek
Hyo Jung Cho
Soon Sun Kim
Ji Eun Han
Jae Youn Cheong
Min Seo Shim
Kyoung Jin Lee
Tae Hyun Kim
Chi Yong Ban
Charmgil Hong
|
¿ø¹®¼ö·Ïó(Citation) |
VOL 49 NO. 02 PP. 1676 ~ 1678 (2022. 12) |
Çѱ۳»¿ë (Korean Abstract) |
|
¿µ¹®³»¿ë (English Abstract) |
This paper aims at predicting drug-induced liver injury (DILI). Many natural or manufactured chemical compounds entering our body are filtered and metabolized within our liver system. As such the liver is susceptible to injuries, especially when one is hospitalized. This work uses the in-hospital observations of patients stored in electronic health records (EHRs) and predicts the potential DILI occurrences in the future by building an effective representation of the patients using Bi-Directional Long-Short Term Memory (BI-LSTM). With 0.8771 AUROC and 0.2289 AUPRC, the proposed method is shown to properly predict the occurrence of DILI. |
Å°¿öµå(Keyword) |
|
ÆÄÀÏ÷ºÎ |
PDF ´Ù¿î·Îµå
|