Predicting Early Neonatal Sepsis using Neural Networks and Other Classifiers

The aim of this paper is to develop an Artificial Neural Network (ANN) model which will act as a classifier for detecting early-onset sepsis in neonates. Neonatal sepsis is a major concern for maternal and neonatal health which remains a global problem with little progress made despite major efforts. This study aims to improve the existing work on an imbalanced dataset from Crecer’s Hospital Centre in Cartagena-Colombia. Standard oversampling and under sampling are used to balance the dataset. A set of input and hyper-parameter different from the existing work is fed to the ANN. Results of this study show a true positive rate (recall) of 98.4%, true negative rate (specificity) of 98.1%, and a precision of 96.8%. An area under the curve (AUC) of 99.8% has been achieved. With an accuracy of 98.2%, this neural network succeeds in improving on the previous model for this data.

[1]  Yaxin Bi,et al.  KNN Model-Based Approach in Classification , 2003, OTM.

[2]  Geoffrey E. Hinton,et al.  Visualizing Data using t-SNE , 2008 .

[3]  Sanyam Shukla,et al.  Dynamic selection of normalization techniques using data complexity measures , 2018, Expert Syst. Appl..

[4]  Chee Kheong Siew,et al.  Extreme learning machine: Theory and applications , 2006, Neurocomputing.

[5]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[6]  Imas Sukaesih Sitanggang,et al.  Parameter Tuning in Random Forest Based on Grid Search Method for Gender Classification Based on Voice Frequency , 2017 .

[7]  Christopher P. Bonafide,et al.  Machine learning models for early sepsis recognition in the neonatal intensive care unit using readily available electronic health record data , 2019, PloS one.

[8]  X.,et al.  Data Mining for Direct Marketing : Problems and , 1998 .

[9]  Hye Jin Kam,et al.  Learning representations for the early detection of sepsis with deep neural networks , 2017, Comput. Biol. Medicine.

[10]  Stephen Marshall,et al.  Activation Functions: Comparison of trends in Practice and Research for Deep Learning , 2018, ArXiv.

[11]  Mehryar Mohri,et al.  L2 Regularization for Learning Kernels , 2009, UAI.

[12]  T. Minka.  A comparison of numerical optimizers for logistic regression , 2004 .

[13]  Dmitry Yarotsky,et al.  Error bounds for approximations with deep ReLU networks , 2016, Neural Networks.

[14]  Uli K. Chettipally,et al.  Prediction of Sepsis in the Intensive Care Unit With Minimal Electronic Health Record Data: A Machine Learning Approach , 2016, JMIR medical informatics.

[15]  M. P. Griffin,et al.  Heart Rate Characteristics and Laboratory Tests in Neonatal Sepsis , 2005, Pediatrics.

[16]  Robert Steele,et al.  Implementation of a bundle of quality indicators for the early management of severe sepsis and septic shock is associated with decreased mortality* , 2007, Critical care medicine.

[17]  Nathan Srebro,et al.  The Marginal Value of Adaptive Gradient Methods in Machine Learning , 2017, NIPS.

[18]  Vicente García-Díaz,et al.  A neural network approach to predict early neonatal sepsis , 2019, Comput. Electr. Eng..

[19]  Xu Sun,et al.  Adaptive Gradient Methods with Dynamic Bound of Learning Rate , 2019, ICLR.

[20]  Steven Horng,et al.  Creating an automated trigger for sepsis clinical decision support at emergency department triage using machine learning , 2017, PloS one.

[21]  Christopher W. Barton,et al.  A computational approach to early sepsis detection , 2016, Comput. Biol. Medicine.

[22]  P. Adatara,et al.  Risk Factors Associated with Neonatal Sepsis: A Case Study at a Specialist Hospital in Ghana , 2019, TheScientificWorldJournal.

[23]  Jian Sun,et al.  Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet Classification , 2015, 2015 IEEE International Conference on Computer Vision (ICCV).

[24]  Constantin F. Aliferis,et al.  Medical decision support using machine learning for early detection of late-onset neonatal sepsis , 2014, J. Am. Medical Informatics Assoc..

[25]  Francisco Herrera,et al.  SMOTE for Learning from Imbalanced Data: Progress and Challenges, Marking the 15-year Anniversary , 2018, J. Artif. Intell. Res..