Benchmark of Deep Learning Models on Large Healthcare MIMIC Datasets (2017-10-23T00:00:00.000000Z)

TL;DR

This paper benchmarked the performance of the deep learning models with respect to the state-of-the-art machine learning models and prognostic scoring systems on publicly available healthcare datasets for several clinical prediction tasks such as mortality prediction, length of stay prediction, and ICD-9 code group prediction.

Abstract

Deep learning models (aka Deep Neural Networks) have revolutionized many fields including computer vision, natural language processing, speech recognition, and is being increasingly used in clinical healthcare applications. However, few works exist which have benchmarked the performance of the deep learning models with respect to the state-of-the-art machine learning models and prognostic scoring systems on publicly available healthcare datasets. In this paper, we present the benchmarking results for several clinical prediction tasks such as mortality prediction, length of stay prediction, and ICD-9 code group prediction using Deep Learning models, ensemble of machine learning models (Super Learner algorithm), SAPS II and SOFA scores. We used the Medical Information Mart for Intensive Care III (MIMIC-III) (v1.4) publicly available dataset, which includes all patients admitted to an ICU at the Beth Israel Deaconess Medical Center from 2001 to 2012, for the benchmarking tasks. Our results show that deep learning models consistently outperform all the other approaches especially when the `raw' clinical time series data is used as input features to the models.

Authors

Chuizheng Meng

3 papers

Yan Liu

7 papers

Zhengping Che

3 papers

TL;DR

Abstract

Authors

References62 items

Reproducibility in critical care: a mortality prediction case study

Multitask learning and benchmarking with clinical time series data

Patient-Specific Predictive Modeling Using Random Forests: An Observational Study for the Critically Ill

Customization of a Severity of Illness Score Using Local Electronic Medical Record Data

Variational Recurrent Adversarial Deep Domain Adaptation

Identifiable Phenotyping using Constrained Non-Negative Matrix Factorization

A computational approach to mortality prediction of alcohol use disorder inpatients

Prediction using patient comparison vs. modeling: A case study for mortality prediction

Recurrent Neural Networks for Multivariate Time Series with Missing Values

MIMIC-III, a freely accessible critical care database

Predicting ICU Mortality Risk by Grouping Temporal Trends from a Multivariate Panel of Physiologic Measurements

Distilling Knowledge from Deep Networks with Applications to Healthcare Domain

Learning to Diagnose with LSTM Recurrent Neural Networks

A Neural Network Based Model for Predicting Psychological Conditions

Dynamically Modeling Patient's Health State from Electronic Medical Records: A Time Series Approach

Deep Computational Phenotyping

Personalized Mortality Prediction Driven by Electronic Medical Data and a Patient Similarity Metric

A Multivariate Timeseries Modeling Approach to Severity of Illness Assessment and Forecasting in ICU with Sparse, Heterogeneous Clinical Data

PD Disease State Assessment in Naturalistic Environments Using Deep Learning

Empirical Evaluation of Gated Recurrent Neural Networks on Sequence Modeling

Going deeper with convolutions

On the Properties of Neural Machine Translation: Encoder–Decoder Approaches

Unfolding physiological state: mortality modelling in intensive care units

Caffe: Convolutional Architecture for Fast Feature Embedding

Distributed Representations of Words and Phrases and their Compositionality

Computational Phenotype Discovery Using Unsupervised Feature Learning over Noisy, Sparse, and Irregular Clinical Data

Linguistic Regularities in Continuous Space Word Representations

Multimodal learning with deep Boltzmann machines

ImageNet classification with deep convolutional neural networks

Theano: new features and speed improvements

Predicting in-hospital mortality of ICU patients: The PhysioNet/Computing in cardiology challenge 2012

A Database-driven Decision Support System: Customized Mortality Prediction

Representation Learning: A Review and New Perspectives

Joint Learning of Words and Meaning Representations for Open-Text Semantic Parsing

Context-Dependent Pre-Trained Deep Neural Networks for Large-Vocabulary Speech Recognition

Open-access MIMIC-II database for intensive care research

A Comparison of Intensive Care Unit Mortality Prediction Models through the Use of Data Mining Techniques

Empirical Evaluation and Combination of Advanced Language Modeling Techniques

Phone Recognition with the Mean-Covariance Restricted Boltzmann Machine

Rectified Linear Units Improve Restricted Boltzmann Machines

Exploring Strategies for Training Deep Neural Networks

Artificial intelligence applications in the intensive care unit

Long Short-Term Memory

The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure

Prediction of outcome in critically ill patients using artificial neural network synthesised by genetic algorithm

Using the Future to Sort Out the Present: Rankprop and Multitask Learning for Medical Risk Evaluation

Use of a neural network as a predictive instrument for length of stay in the intensive care unit following cardiac surgery.

Artificial intelligence in medicine

Multilayer feedforward networks are universal approximators

A Learning Algorithm for Continually Running Fully Recurrent Neural Networks

APACHE—acute physiology and chronic health evaluation: a physiologically based classification system

Mortality Prediction in the ICU Based on MIMIC-II Results from the Super ICU Learner Algorithm (SICULA) Project

Interpretable Topic Features for Post-ICU Mortality Prediction

Interpretable Deep Models for ICU Outcome Prediction

Deep Learning

Keras: Theano-based deep learning library

Risk Stratification of ICU Patients Using Topic Models Inferred from Unstructured Progress Notes

Prognostic Physiology: Modeling Patient Severity in Intensive Care Units Using Radial Domain Folding

Super learner in prediction, U.C. Berkeley Division of

Super learner, Statistical applications in genetics and molecular biology

A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study.

Modeling mortality in the intensive care unit: comparing the performance of a back-propagation, associative-learning neural network with multivariate logistic regression.

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names