Online citations, reference lists, and bibliographies.

A Universal Deep Learning Approach For Modeling The Flow Of Patients Under Different Severities

Shancheng Jiang, Kwai-Sang Chin, Kwok-Leung Tsui
Published 2018 · Computer Science, Medicine
Cite This
Download PDF
Analyze on Scholarcy
Share
BACKGROUND AND OBJECTIVE The Accident and Emergency Department (A&ED) is the frontline for providing emergency care in hospitals. Unfortunately, relative A&ED resources have failed to keep up with continuously increasing demand in recent years, which leads to overcrowding in A&ED. Knowing the fluctuation of patient arrival volume in advance is a significant premise to relieve this pressure. Based on this motivation, the objective of this study is to explore an integrated framework with high accuracy for predicting A&ED patient flow under different triage levels, by combining a novel feature selection process with deep neural networks. METHODS Administrative data is collected from an actual A&ED and categorized into five groups based on different triage levels. A genetic algorithm (GA)-based feature selection algorithm is improved and implemented as a pre-processing step for this time-series prediction problem, in order to explore key features affecting patient flow. In our improved GA, a fitness-based crossover is proposed to maintain the joint information of multiple features during iterative process, instead of traditional point-based crossover. Deep neural networks (DNN) is employed as the prediction model to utilize their universal adaptability and high flexibility. In the model-training process, the learning algorithm is well-configured based on a parallel stochastic gradient descent algorithm. Two effective regularization strategies are integrated in one DNN framework to avoid overfitting. All introduced hyper-parameters are optimized efficiently by grid-search in one pass. RESULTS As for feature selection, our improved GA-based feature selection algorithm has outperformed a typical GA and four state-of-the-art feature selection algorithms (mRMR, SAFS, VIFR, and CFR). As for the prediction accuracy of proposed integrated framework, compared with other frequently used statistical models (GLM, seasonal-ARIMA, ARIMAX, and ANN) and modern machine models (SVM-RBF, SVM-linear, RF, and R-LASSO), the proposed integrated "DNN-I-GA" framework achieves higher prediction accuracy on both MAPE and RMSE metrics in pairwise comparisons. CONCLUSIONS The contribution of our study is two-fold. Theoretically, the traditional GA-based feature selection process is improved to have less hyper-parameters and higher efficiency, and the joint information of multiple features is maintained by fitness-based crossover operator. The universal property of DNN is further enhanced by merging different regularization strategies. Practically, features selected by our improved GA can be used to acquire an underlying relationship between patient flows and input features. Predictive values are significant indicators of patients' demand and can be used by A&ED managers to make resource planning and allocation. High accuracy achieved by the present framework in different cases enhances the reliability of downstream decision makings.
This paper references
10.1109/TNNLS.2014.2308519
Learning Deep and Wide: A Spectral Method for Learning Deep Networks
Ling Shao (2014)
10.1016/j.neucom.2015.09.116
Deep learning for visual understanding: A review
Yanming Guo (2016)
10.1002/9780470172261
Genetic algorithms and engineering optimization
M. Gen (1999)
10.1016/j.cmpb.2016.02.013
Classification of lung sounds using higher-order statistics: A divide-and-conquer approach
Raphael Naves (2016)
10.1162/153244303322753616
An Introduction to Variable and Feature Selection
Isabelle Guyon (2003)
10.1016/j.cmpb.2014.08.006
Scheduling prioritized patients in emergency department laboratories
Ali Azadeh (2014)
10.1016/j.cmpb.2016.09.018
Detection of exudates in fundus photographs using deep neural networks and anatomical landmark detection fusion
Pavle Prentasic (2016)
ADADELTA: An Adaptive Learning Rate Method
Matthew D. Zeiler (2012)
10.1023/A:1022627411411
Support-Vector Networks
Corinna Cortes (2004)
10.1109/TMM.2016.2591508
Learning Cascaded Deep Auto-Encoder Networks for Face Alignment
Renliang Weng (2016)
10.1109/TSMCA.2012.2217319
A Three-Level Strategy for the Design and Performance Evaluation of Hospital Departments
Maria Pia Fanti (2013)
10.1016/j.cmpb.2013.12.005
Hybridization between multi-objective genetic algorithm and support vector machine for feature selection in walker-assisted gait
Maria M. Martins (2014)
10.1007/978-3-7908-2604-3_16
Large-Scale Machine Learning with Stochastic Gradient Descent
Léon Bottou (2010)
10.1111/acem.12182
Forecasting daily emergency department visits using calendar variables and ambient temperature readings.
Izabel Marcilio (2013)
10.1016/S0893-6080(03)00138-2
The general inefficiency of batch training for gradient descent learning
D. Randall Wilson (2003)
10.1109/TIP.2017.2678163
Deep Hashing for Scalable Image Search
Jiwen Lu (2017)
10.1109/JBHI.2015.2404829
Predicting Asthma-Related Emergency Department Visits Using Big Data
Sudha Ram (2015)
10.1016/j.ajem.2014.03.011
A retrospective analysis of the utility of an artificial neural network to predict ED volume.
Nathan Benjamin Menke (2014)
10.1109/CVPR.2007.383157
Unsupervised Learning of Invariant Feature Hierarchies with Applications to Object Recognition
Marc'Aurelio Ranzato (2007)
10.1016/j.dss.2012.08.002
Rescheduling of elective patients upon the arrival of emergency patients
Ergin Erdem (2012)
10.1080/0740817X.2011.609875
Reducing emergency department waiting times by adjusting work shifts considering patient visits to multiple care providers
David Sinreich (2012)
10.1038/nature14539
Deep Learning
Ian G Goodfellow (2015)
10.1016/j.dss.2012.12.019
Modeling daily patient arrivals at Emergency Department and quantifying the relative importance of contributing variables using artificial neural network
M. Xu (2013)
10.18637/JSS.V053.I04
GA: A Package for Genetic Algorithms in R
Luca Scrucca (2013)
10.1016/j.dss.2010.08.028
Combining multiple feature selection methods for stock prediction: Union, intersection, and multi-intersection approaches
Chih-Fong Tsai (2010)
10.1109/CVPR.2016.133
Learning Compact Binary Descriptors with Unsupervised Deep Neural Networks
Kevin Lin (2016)
10.1016/j.cmpb.2011.02.010
Optimization of scheduling patient appointments in clinics using a novel modelling technique of patient arrival
Guangfu Tai (2012)
10.1111/j.1553-2712.2008.00083.x
The challenge of predicting demand for emergency department services.
Melissa Lee McCarthy (2008)
10.1016/j.compeleceng.2013.11.024
A survey on feature selection methods
Girish Chandrashekar (2014)
10.18637/JSS.V053.I06
Interactive network exploration with Orange
Miha Štajdohar (2013)
10.1136/emj.2010.103531
Predicting emergency department admissions
Justin R. Boyle (2011)
10.1111/j.1553-2712.2009.00356.x
Forecasting models of emergency department crowding.
Lisa M. Schweigler (2009)
10.1016/j.infsof.2010.05.009
GA-based method for feature selection and parameters optimization for machine learning regression applied to software effort estimation
Adriano Lorena Inácio de Oliveira (2010)
10.1016/j.cmpb.2015.06.006
A medical cost estimation with fuzzy neural network of acute hepatitis patients in emergency room
R. J. Kuo (2015)
Improving neural networks by preventing co-adaptation of feature detectors
Geoffrey E. Hinton (2012)
Hogwild: A Lock-Free Approach to Parallelizing Stochastic Gradient Descent
Benjamin Recht (2011)
10.1016/j.eswa.2005.09.024
A GA-based feature selection and parameters optimizationfor support vector machines
C. Huang (2006)
10.1186/1471-227X-9-1
Forecasting daily attendances at an emergency department to aid resource planning
Yan Sun (2009)
10.1016/j.eswa.2005.11.027
A hybrid SARIMA and support vector machines in forecasting the production values of the machinery industry in Taiwan
Kuan-Yu Chen (2007)
10.1136/emj.2006.045310
Effects of ambient temperature on volume, specialty composition and triage levels of emergency department visits
Chia-Chun Tai (2007)
10.1038/srep10312
Application of high-dimensional feature selection: evaluation for genomic prediction in man
Mairead Lesley Bermingham (2015)
10.1016/j.neunet.2014.09.003
Deep learning in neural networks: An overview
Jürgen Schmidhuber (2015)
10.1111/j.1553-2712.2007.00032.x
Forecasting daily patient volumes in the emergency department.
Spencer S. Jones (2008)
10.1016/S0169-7439(98)00051-3
Genetic algorithms applied to feature selection in PLS regression: how and when to use them
Riccardo Leardi (1998)
10.1023/A:1018054314350
Bagging Predictors
L. Breiman (1996)
Dropout: a simple way to prevent neural networks from overfitting
Nitish Srivastava (2014)
10.1016/j.cmpb.2010.03.015
Long-term staff scheduling with regular temporal distribution
Rafael C. Carrasco (2010)
10.1561/2200000006
Learning Deep Architectures for AI
Yoshua Bengio (2007)
10.2196/MHEALTH.6562
Sleep Quality Prediction From Wearable Data Using Deep Learning
Aarti Sathyanarayana (2016)
10.1023/A:1010933404324
Random Forests
Leo Breiman (2004)



This paper is referenced by
10.1016/J.INFRARED.2018.12.001
ALK positive lung cancer identification and targeted drugs evaluation using microscopic hyperspectral imaging technique
Jing Song (2019)
10.4018/IJCCP.2018070103
The Role of Artificial Intelligence in Clinical Decision Support Systems and a Classification Framework
Steven Walczak (2018)
10.3233/MAS-180448
Shorter waiting time, better emergency healthcare: Forecasting Stockholm's emergency department visits
Andreas Ekström (2018)
10.1016/j.zemedi.2018.11.002
An overview of deep learning in medical imaging focusing on MRI
Alexander Selvikvåg Lundervold (2019)
10.1177/1460458219871780
A machine learning–based 1-year mortality prediction model after hospital discharge for clinical patients with acute coronary syndrome
Syed Waseem Abbas Sherazi (2019)
10.1016/J.IJMEDINF.2019.05.011
Short and Long term predictions of Hospital emergency department attendances
Tahseen Ahmed Jilani (2019)
10.1080/20476965.2018.1547348
An exhaustive review and analysis on applications of statistical forecasting in hospital emergency departments
Muhammet Gul (2018)
10.3390/ijerph17082664
Methodological Approaches to Support Process Improvement in Emergency Departments: A Systematic Review
Miguel Angel Ortíz-Barrios (2020)
10.1016/S0169-2607(17)31508-0
Improving healthcare management with data science
Hung-Wen Chiu (2018)
10.1002/hpm.2771
A comprehensive modelling framework to forecast the demand for all hospital services.
Muhammed Ordu (2019)
10.1186/s13104-019-4393-y
Comparative evaluation of time series models for predicting influenza outbreaks: application of influenza-like illness data from sentinel sites of healthcare centers in Iran
L. Tapak (2019)
Semantic Scholar Logo Some data provided by SemanticScholar