(주) 뷰노

(주) 뷰노

Submit Inquiries
KR EN
Publications
BACK TO LIST
Critical Care Medicine

Graph Convolutional Networks-based Noisy Data Imputation in Electronic Health Record

  • Nov. 2020
  • by Byeong Tak Lee et. al.

Objectives: 

A deep learning-based early warning system is proposed to predict sepsis prior to its onset.

Design: 

A novel algorithm was devised to detect sepsis 6 hours prior to its onset based on electronic medical records.

Setting: 

Retrospective cohorts from three separate hospitals are used in this study. Sepsis onset was defined based on Sepsis-3. Algorithms are evaluated based on the score function used in the Physionet Challenge 2019.Patients: Over 60,000 ICU patients with 40 clinical variables (vital signs, laboratory results) for each hour of a patient’s ICU stay were used.

Interventions: 

None.

Measurements and Main Results: 

The proposed algorithm predicted the onset of sepsis in the preceding n hours (where n = 4, 6, 8, or 12). Furthermore, the proposed method compared how many sepsis patients can be predicted in a short time with other methods. To interpret a given result in a clinical perspective, the relationship between input variables and the probability of the proposed method were presented. The proposed method achieved superior results (area under the receiver operating characteristic curve, area under the precision-recall curve, and score) and predicted more sepsis patients in advance. In official phase, the proposed method showed the utility score of –0.101, area under the receiver operating characteristic curve 0.782, area under the precision-recall curve 0.041, accuracy 0.786, and F-measure 0.046.

Conclusions: 

Using Physionet Challenge 2019, the proposed method can accurately and early predict the onset of sepsis. The proposed method can be a practical early warning system in the environment of real hospitals.

Link

https://journals.lww.com/ccmjournal/Abstract/2020/11000/Graph_Convolutional_Networks_Based_Noisy_Data.40.aspx

Author

Byeong Tak Lee, O-Yeon Kwon, Hyunho Park, Kyung-Jae Cho, Joon-Myoung Kwon, and Yeha Lee

Tag in

#biosignal

Previous
Development and Validation of a Deep Learning–Based Automatic Brain Segmentation and Classification Algorithm for Alzheimer Disease Using 3D T1-Weighted Volumetric Images 11. 30. 2020
Next
Evolved Speech-Transformer: Applying Neural Architecture Search to End-to-End Automatic Speech Recognition 11. 30. 2020
List

We use cookies or similar technologies as specified in our privacy policy to enhance your experience.

If you would like to learn more about how we use cookies, click "Privacy Policy".