Deep Patient: An Unsupervised Representation to Predict the Future of Patients from the Electronic Health Records
By Riccardo Miotto, Li Li, Brian A. Kidd, Joel T. Dudley. 2016
1,635 citations recorded by the source corpus.
Secondary use of electronic health records (EHRs) promises to advance clinical research and better inform clinical decision making. Challenges in summarizing and representing patient data prevent widespread practice of predictive modeling using EHRs. Here we present a novel unsupervised deep feature learning method to derive a general-purpose patient representation from EHR data that facilitates clinical predictive modeling. In particular, a three-layer stack of denoising autoencoders was used to capture hierarchical regularities and dependencies in the aggregated EHRs of about 700,000 patients from the Mount Sinai data warehouse. The result is a representation we name “deep patient”. We evaluated this representation as broadly predictive of health states by assessing the probability of patients to develop various diseases. We performed evaluation using 76,214 test patients comprising 78 diseases from diverse clinical domains and temporal windows. Our results significantly outperformed those achieved using representations based on raw EHR data and alternative feature learning strategies. Prediction performance for severe diabetes, schizophrenia, and various cancers were among the top performing. These findings indicate that deep learning applied to EHRs can derive patient representations that offer improved clinical predictions, and could provide a machine learning framework for augmenting clinical decision systems.
Subfields: Artificial Intelligence, Health Information Management, Molecular Biology, Representation & Self-Supervised Learning, Medical & Health AI
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
- Mohit Bansal
- Yoshua Bengio
- Quoc V. Le
- Jianfeng Gao
- Percy Liang
- Ruslan Salakhutdinov
- Noah A. Smith
- Sergey Levine
In Health Information Management
In Molecular Biology
- Fei Wang
- Jimeng Sun
- Edward Choi
- Wen Torng
- Casey S. Greene
- Russ B. Altman
- Amr M. Alexandari
- Avanti Shrikumar
In Representation & Self-Supervised Learning
In Medical & Health AI
Citations in the corpus
- Beyond Sparsity: Tree Regularization of Deep Models for Interpretability (2017, background): “(Miotto et al. 2016; Choi et al. 2016; Che et al. 2015)), the clinical community remains skeptical of machine learning systems”
- Opportunities and obstacles for deep learning in biology and medicine (2017, methodology): “[82] used a deep denoising autoencoder architecture applied to the number and co-occurrence of clinical events (DeepPatient) to learn a representation of patients.”
- Doctor AI: Predicting Clinical Events via Recurrent Neural Networks (2015, neutral): “We make the source code of Doctor AI publicly available at https://github.com/mp2893/doctorai .”
- GRAM: Graph-based Attention Model for Healthcare Representation Learning (2016, background): “3098126 networks (RNN) [7, 9, 10, 25], convolutional neural networks (CNN) [30] or stacked denoising autoencoders (SDA) [6, 29], demonstrating signi cant performance enhancement for diverse prediction tasks.”
- Scalable and accurate deep learning with electronic health records (2018, methodology): “In a wellknown work, investigators used auto-encoders to predict a specific set of diagnoses.(29) Subsequent work extended this approach by modeling the temporal sequence of events that occurred in a patient’s record, which may enhance accuracy in scenarios that depend on the order of events, with c”
- Deep Learning for Health Informatics (2017, background): “[102] showed that a stack of denoising autoencoders can be used to automatically infer features from a large-scale EHR database and represent patients with-”
- Leveraging uncertainty information from deep neural networks for disease detection (2016, methodology): “We then approximate the integral by Monte Carlo sampling and compute the predictive mean (to be used for a final prediction 302 on a test image): 303 μpred ≈ 1 T T ∑ t=1 p(y|x,θ(ω̂t)) (7)”
- Mining Electronic Health Records: A Survey (2017, background): “Miotto et al. [Miotto et al. 2016] proposed a novel unsupervised feature learning algorithm using three layer stack of autoencoders using EHRs that faciliate predictive modeling for various diseases such as T2DM, cancer and schizophrenia.”
- Identify Susceptible Locations in Medical Records via Adversarial Attacks on Deep Predictive Models (2018, methodology): “odern machine learning techniques have also demonstrated great potentials in clinical informatics [36]. For example, deep learning has been used to learn effective representations for patient records [14, 26] to support disease phenotyping [8] and conduct predictive modeling [9, 30, 33]. A recent st”
- Learning Tasks for Multitask Learning: Heterogenous Patient Populations in the ICU (2018, methodology): “ our task because of the complex temporal dependencies in physiological time-series. Autoencoders have been used to learn compact representations of patient state from multi-modal timeseries EHR data [27, 40]. We use one LSTM layer in the encoder and one in the decoder. The embedded representation i”
- Two-stage Federated Phenotyping and Patient Representation Learning (2019, background): “Patient representation learning has been done in a supervised (Dligach and Miller, 2018) and unsupervised (Miotto et al., 2016) setting.”
- Using clinical Natural Language Processing for health outcomes research: Overview and actionable suggestions for future advances (2018, neutral): “There are cases in which an NLP method might be used to process all of the relevant text associated with a single patient, over time, in order to directly predict a single outcome; for example, all past text (and other information) could be used in order to assign a diagnosis code [64].”