|
KMID : 0311120220630070640
|
|
Yonsei Medical Journal
2022 Volume.63 No. 7 p.640 ~ p.647
|
|
|
Application of Machine Learning Approaches to Predict Postnatal Growth Failure in Very Low Birth Weight Infants
|
|
Han Jung-Ho
Yoon So-Jin
Lee Hye-Sun
Park Go-Eun
Lim Joo-Hee
Shin Jeong-Eun
Eun Ho-Seon
Park Min-Soo
Lee Soon-Min
|
|
|
Abstract
|
|
|
Purpose: The aims of the study were to develop and evaluate a machine learning model with which to predict postnatal growthfailure (PGF) among very low birth weight (VLBW) infants.
Materials and Methods: Of 10425 VLBW infants registered in the Korean Neonatal Network between 2013 and 2017, 7954 infantswere included. PGF was defined as a decrease in Z score >1.28 at discharge, compared to that at birth. Six metrics [area under the re ceiver operating characteristic curve (AUROC), accuracy, precision, sensitivity, specificity, and F1 score] were obtained at five timepoints (at birth, 7 days, 14 days, 28 days after birth, and at discharge). Machine learning models were built using four different tech niques [extreme gradient boosting (XGB), random forest, support vector machine, and convolutional neural network] to compareagainst the conventional multiple logistic regression (MLR) model.
Results: The XGB algorithm showed the best performance with all six metrics across the board. When compared with MLR, XGBshowed a significantly higher AUROC (p=0.03) for Day 7, which was the primary performance metric. Using optimal cut-off points,for Day 7, XGB still showed better performances in terms of AUROC (0.74), accuracy (0.68), and F1 score (0.67). AUROC valuesseemed to increase slightly from birth to 7 days after birth with significance, almost reaching a plateau after 7 days after birth.
Conclusion: We have shown the possibility of predicting PGF through machine learning algorithms, especially XGB. Such mod els may help neonatologists in the early diagnosis of high-risk infants for PGF for early intervention.
|
|
|
KEYWORD
|
|
|
Growth failure, very low birth weight infants, machine learning, prediction, neonatal intensive care unit
|
|
|
FullTexts / Linksout information
|
|
|
|
|
Listed journal information
|
|
   
|
|