Next-generation phenotyping of electronic health records

By George Hripcsak, David J. Albers. 2012

743 citations recorded by the source corpus.

The national adoption of electronic health records (EHR) promises to make an unprecedented amount of data available for clinical research, but the data are complex, inaccurate, and frequently missing, and the record reflects complex processes aside from the patient's physiological state. We believe that the path forward requires studying the EHR as an object of interest in itself, and that new models, learning from data, and collaboration will lead to efficient use of the valuable information currently locked in health records.

Subfields: Artificial Intelligence, Health Information Management, Molecular Biology

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

In Health Information Management

In Molecular Biology

Citations in the corpus

Natural Language Processing for EHR-Based Computational Phenotyping (2018, background): “For example, ICD-9 codes are mainly recorded for administrative purposes and are influenced by billing requirements and avoidance of liability [12, 13].”
Large-Scale Discovery of Disease-Disease and Disease-Gene Associations (2016, background): “In addition, to create more scalable, effective tools, recent approaches distinct from networks have focused on the development of data-driven phenotyping with minimal manual input and rigorous evaluation procedures 3,30,31 .”
Deep learning for healthcare: review, opportunities and challenges (2018, neutral): “In fact, there remain many challenges in making full use of the biomedical data, owing to their high-dimensionality, heterogeneity, temporal dependency, sparsity and irregularity [13–15].”
Comparing Rule-Based and Deep Learning Models for Patient Phenotyping (2017, methodology): “Extracted entities are filtered to only include concepts related to the patient phenotype under investigation and either used as features for a model that predicts whether the patient fits the phenotype, or as input for rule-based algorithms.(18,38,48) Liao et al.”
Deep Survival Analysis (2016, background): “First, EHR data is usually high-dimensional and very sparse (Hripcsak and Albers, 2013; Pivovarov et al., 2014).”
Limestone: High-throughput candidate phenotype generation via tensor factorization (2014, methodology): “Data mining and machine learning tools have been utilized to automate the phenotype generation process [24,36,37,26].”
Using clinical Natural Language Processing for health outcomes research: Overview and actionable suggestions for future advances (2018, neutral): “While there have been calls for increased structuring of health records, these seem to be mainly driven by convenience issues for researchers or administrators (i.e. ease of access to pre-structured data) rather than the preferences of the clinical staff actually entering data [39].”
Caveats for the Use of Operational Electronic Health Record Data in Comparative Effectiveness Research (2013, neutral): “Another analyses have highlighted the potential value but also the cautions of using EHR for research purposes.(4,27) In this paper, we describe the caveats of using operational EHR data for CER and provide recommendations for moving forward.”
Sparse Multi-Output Gaussian Processes for Medical Time Series Prediction (2017, background): “Large-scale collections of electronic health records (EHRs) are becoming useful for understanding disease progress, early diagnosis, and personalized treatments for many clinical diseases (Murdoch and Detsky, 2013; Hripcsak and Albers, 2013; Ghassemi et al., 2015a).”
Identifying and Mitigating Biases in EHR Laboratory Tests (2014, neutral): “Automating feature selection from EHR variables is a difficult task, as the EHR is an inherently biased data source: EHR data are collected with the primary goal of delivering and documenting patient care, not with the primary goal of creating a curated research dataset [4] [5].”
Generative Adversarial Networks for Electronic Health Records: A Framework for Exploring and Evaluating Methods for Predicting Drug-Induced Laboratory Test Trajectories (2017, background): “EHR data are known to be complex due to their multi-modality [6], mixing categorical and continuous data with semi-structured and free text medical notes.”
Learning Optimal Individualized Treatment Rules from Electronic Health Record Data (2016, background): “Obtain (cid:101) R i = R i − (cid:98) R i for each subject, and ﬁt a weighted classiﬁcation tree to estimate the decision function f , where the weights are | (cid:101) R i / (cid:98) π ( A i , H i ) | , and the class labels are A i sign ( (cid:101) R i ) .”
Comparing deep learning and concept extraction based methods for patient phenotyping from clinical narratives (2018, methodology): “A popular approach to patient phenotyping using NLP is based on extracting medical phrases from texts and using them as input to build a predictive model [19].”
DeepCare: A Deep Dynamic Memory Model for Predictive Medicine (2016, background): “Like most applications of machine learning, the bottleneck here is manual feature engineering due to the complexity of the data [22][21].”