Machine Learning and Decision Support in Critical Care

By Alistair E. W. Johnson, Marzyeh Ghassemi, Shamim Nemati, Katherine E. Niehaus, David A. Clifton, G. Clifford. 2016

408 citations recorded by the source corpus.

Clinical data management systems typically provide caregiver teams with useful information, derived from large, sometimes highly heterogeneous, data sources that are often changing dynamically. Over the last decade there has been a significant surge in interest in using these data sources, from simply reusing the standard clinical databases for event prediction or decision support, to including dynamic and patient-specific information into clinical monitoring and prediction problems. However, in most cases, commercial clinical databases have been designed to document clinical activity for reporting, liability, and billing reasons, rather than for developing new algorithms. With increasing excitement surrounding “secondary use of medical records” and “Big Data” analytics, it is important to understand the limitations of current databases and what needs to change in order to enter an era of “precision medicine.” This review article covers many of the issues involved in the collection and preprocessing of critical care data. The three challenges in critical care are considered: compartmentalization, corruption, and complexity. A range of applications addressing these issues are covered, including the modernization of static acuity scoring; online patient tracking; personalized prediction and risk assessment; artifact detection; state estimation; and incorporation of multimodal data sources such as genomic and free text data.

Subfields: Artificial Intelligence

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

Citations in the corpus

Reinforcement Learning in Healthcare: A Survey (2019, methodology): “However, the inherent 3C (Compartmentalization, Corruption, and Complexity) features indicate that critical care data are usually noisy, biased and incomplete [5].”
Deep Learning for Health Informatics (2017, background): “For patients in critical care, continuous monitoring of vital signs, such as blood pressure, respiration rate, and body temperature, are important for improving treatment outcomes by closely analyzing the patient’s condition [137].”
Inverse reinforcement learning for intelligent mechanical ventilation and sedative dosing in intensive care units (2019, background): “…the development in ubiquitous monitoring techniques, a plethora of ICU data has been generated in a variety of formats such as free-text clinical notes, images, physiological waveforms, and vital sign time series, enable optimal diagnose, treat and mortality prediction of a patient in ICUs [15].”
Deep Self-Organization: Interpretable Discrete Representation Learning on Time Series (2018, background): “Clinicians, for instance, might need to visualize and manually process the trajectory of a patient in order to decide on their treatment [17].”
A Non-parametric Learning Method for Confidently Estimating Patient's Clinical State and Dynamics (2016, methodology): “This is a vital step in medical applications since the results of the non-parametric Bayesian inference algorithm are sensitive to the selected hyper-parameters [14, 12].”
Predicting Complications in Critical Care Using Heterogeneous Clinical Data (2016, background): “These issues are well described in [5].”
Use of Machine-Learning Approaches to Predict Clinical Deterioration in Critically Ill Patients: A Systematic Review (2017, background): “Although some reviews have reported machine-learning methods in critical care, these reviews have not systematically described the performance outcomes [18,19].”
Machine learning in critical care: state-of-the-art and a sepsis case study (2018, neutral): “A similar attempt by Johnson et al. [23] takes a different route and emphasizes the challenges posed by critical care data themselves.”