Multitask learning and benchmarking with clinical time series data

Abstract

Health care is one of the most exciting frontiers in data mining and machine learning. Successful adoption of electronic health records (EHRs) created an explosion in digital clinical data available for analysis, but progress in machine learning for healthcare research has been difficult to measure because of the absence of publicly available benchmark data sets. To address this problem, we propose four clinical prediction benchmarks using data derived from the publicly available Medical Information Mart for Intensive Care (MIMIC-III) database. These tasks cover a range of clinical problems including modeling risk of mortality, forecasting length of stay, detecting physiologic decline, and phenotype classification. We propose strong linear and neural baselines for all four tasks and evaluate the effect of deep supervision, multitask training and data-specific architectural modifications on the performance of neural models.

This version of the paper adds details about the generation of the benchmark tasks and describes improved neural baselines

References in corpus (4)

• Scalable and accurate deep learning for electronic health records
• CheXNet: Radiologist-Level Pneumonia Detection on Chest X-Rays with Deep Learning
• RETAIN: An Interpretable Predictive Model for Healthcare using Reverse Time Attention Mechanism
• Improving Hospital Mortality Prediction with Medical Named Entities and Multimodal Learning

Cited by in corpus (28)

• Precision Health Data: Requirements, Challenges and Existing Techniques for Data Security and Privacy
• Deep learning for temporal data representation in electronic health records: A systematic review of challenges and methodologies
• Predicting Intensive Care Unit Length of Stay and Mortality Using Patient Vital Signs: Machine Learning Model Development and Validation
• Benchmarking machine learning models on multi-centre eICU critical care dataset
• Annotating Social Determinants of Health Using Active Learning, and Characterizing Determinants Using Neural Event Extraction
• Benchmarking emergency department triage prediction models with machine learning and large public electronic health records
• Synthesize High-dimensional Longitudinal Electronic Health Records via Hierarchical Autoregressive Language Model
• Variationally Regularized Graph-based Representation Learning for Electronic Health Records
• GenHPF: General Healthcare Predictive Framework with Multi-task Multi-source Learning
• MeSIN: Multilevel Selective and Interactive Network for Medication Recommendation
• Cross-Silo Federated Learning for Multi-Tier Networks with Vertical and Horizontal Data Partitioning
• Label Dependent Attention Model for Disease Risk Prediction Using Multimodal Electronic Health Records
• Evaluating the Impact of Social Determinants on Health Prediction in the Intensive Care Unit
• Federated Learning for Clinical Structured Data: A Benchmark Comparison of Engineering and Statistical Approaches
• Accelerating Vertical Federated Learning
• A Multidatabase ExTRaction PipEline (METRE) for Facile Cross Validation in Critical Care Research
• The Missing Indicator Method: From Low to High Dimensions
• Real-time Mortality Prediction Using MIMIC-IV ICU Data Via Boosted Nonparametric Hazards
• FlexCare: Leveraging Cross-Task Synergy for Flexible Multimodal Healthcare Prediction
• MedLens: Improve Mortality Prediction Via Medical Signs Selecting and Regression
• Generalized and Transferable Patient Language Representation for Phenotyping with Limited Data
• COPER: Continuous Patient State Perceiver
• Self Attention with Temporal Prior: Can We Learn More from Arrow of Time?
• LIFE: Learning Individual Features for Multivariate Time Series Prediction with Missing Values
• VBridge: Connecting the Dots Between Features and Data to Explain Healthcare Models
• Efficient Observation Time Window Segmentation for Administrative Data Machine Learning
• Deep Kernel Learning for Mortality Prediction in the Face of Temporal Shift
• Multimodal Routing for Interpretable, Robust, and Auditable Clinical Prediction