272 454

Cited 5 times in

Deep Learning Model for Predicting Intradialytic Hypotension Without Privacy Infringement: A Retrospective Two-Center Study

 Hyung Woo Kim  ;  Seok-Jae Heo  ;  Minseok Kim  ;  Jakyung Lee  ;  Keun Hyung Park  ;  Gongmyung Lee  ;  Song In Baeg  ;  Young Eun Kwon  ;  Hye Min Choi  ;  Dong-Jin Oh  ;  Chung-Mo Nam  ;  Beom Seok Kim 
 FRONTIERS IN MEDICINE, Vol.9 : 878858, 2022-07 
Journal Title
Issue Date
deep learning ; hemodialysis ; intradialytic hypotension ; machine learning ; privacy protection
Objective: Previously developed Intradialytic hypotension (IDH) prediction models utilize clinical variables with potential privacy protection issues. We developed an IDH prediction model using minimal variables, without the risk of privacy infringement.

Methods: Unidentifiable data from 63,640 hemodialysis sessions (26,746 of 79 patients for internal validation, 36,894 of 255 patients for external validation) from two Korean hospital hemodialysis databases were finally analyzed, using three IDH definitions: (1) systolic blood pressure (SBP) nadir <90 mmHg (Nadir90); (2) SBP decrease ≥20 mmHg from baseline (Fall20); and (3) SBP decrease ≥20 mmHg and/or mean arterial pressure decrease ≥10 mmHg (Fall20/MAP10). The developed models use 30 min information to predict an IDH event in the following 10 min window. Area under the receiver operating characteristic curves (AUROCs) and precision-recall curves were used to compare machine learning and deep learning models by logistic regression, XGBoost, and convolutional neural networks.

Results: Among 344,714 segments, 9,154 (2.7%), 134,988 (39.2%), and 149,674 (43.4%) IDH events occurred according to three different IDH definitions (Nadir90, Fall20, and Fall20/MAP10, respectively). Compared with models including logistic regression, random forest, and XGBoost, the deep learning model achieved the best performance in predicting IDH (AUROCs: Nadir90, 0.905; Fall20, 0.864; Fall20/MAP10, 0.863) only using measurements from hemodialysis machine during dialysis session.

Conclusions: The deep learning model performed well only using monitoring measurement of hemodialysis machine in predicting IDH without any personal information that could risk privacy infringement.
Files in This Item:
T202202542.pdf Download
Appears in Collections:
1. College of Medicine (의과대학) > Dept. of Internal Medicine (내과학교실) > 1. Journal Papers
1. College of Medicine (의과대학) > Dept. of Preventive Medicine (예방의학교실) > 1. Journal Papers
Yonsei Authors
Kim, Beom Seok(김범석) ORCID logo https://orcid.org/0000-0002-5732-2583
Kim, Hyung Woo(김형우) ORCID logo https://orcid.org/0000-0002-6305-452X
Nam, Chung Mo(남정모) ORCID logo https://orcid.org/0000-0003-0985-0928
Park, Keun Hyung(박근형)
Lee, Gong Myung(이공명)
Heo, Seok-Jae(허석재) ORCID logo https://orcid.org/0000-0002-8764-7995
사서에게 알리기


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.