Related Experiment Video
Updated: Jan 22, 2026
09:00
TBase - an Integrated Electronic Health Record and Research Database for Kidney Transplant Recipients
Published on: April 13, 2021
5.2K
Scalable and accurate deep learning with electronic health records
Alvin Rajkomar1,2, Eyal Oren1, Kai Chen1
11Google Inc, Mountain View, CA USA.
NPJ Digital Medicine
|July 16, 2019
Summary
Deep learning models accurately predict medical events using raw electronic health record (EHR) data in FHIR format. This approach surpasses traditional models, advancing personalized medicine and healthcare quality without data harmonization.
Area of Science:
- Health Informatics
- Artificial Intelligence in Medicine
- Computational Biology
Background:
- Electronic Health Record (EHR) data holds vast potential for personalized medicine and improved healthcare.
- Traditional predictive modeling relies on labor-intensive extraction of curated variables, discarding significant patient data.
- Standardization of EHR data across institutions is a major challenge for developing robust predictive models.
Purpose of the Study:
- To propose a novel representation of entire raw EHR records using the Fast Healthcare Interoperability Resources (FHIR) format.
- To demonstrate the efficacy of deep learning models with this FHIR-based representation for predicting diverse medical events.
- To validate the approach's accuracy and scalability across multiple healthcare centers without site-specific data harmonization.
Main Methods:
- Developed a sequential data representation for raw EHR records, including clinical notes, based on the FHIR standard.
- Applied deep learning models to this comprehensive EHR data representation.
- Validated the models on de-identified EHR data from 216,221 adult patients across two US academic medical centers.
Main Results:
- Deep learning models achieved high accuracy in predicting in-hospital mortality (AUROC 0.93-0.94), 30-day readmission (AUROC 0.75-0.76), and prolonged length of stay (AUROC 0.85-0.86).
- Models accurately predicted all final discharge diagnoses (frequency-weighted AUROC 0.90).
- Performance consistently outperformed traditional predictive models across all evaluated tasks.
Conclusions:
- The proposed FHIR-based EHR representation combined with deep learning enables accurate, scalable, and harmonized predictive modeling.
- This approach significantly advances the potential of EHR data for personalized medicine and clinical decision support.
- Neural networks can effectively extract relevant clinical information from unstructured patient data within EHRs.
Related Concept Videos
01:11Purpose of Health Records I
1.8K
The vital purpose of health records is to provide a complete and accurate account of a patient's medical history, including communication, diagnostic and therapeutic orders, care planning, research, and quality review.
Here's a breakdown of how health records serve these purposes:
Here's a breakdown of how health records serve these purposes:
1.8K
01:19Purpose of Health Records II
1.4K
Health records serve various essential purposes in the healthcare system. Here are some key purposes:
1.4K
01:24Data Reporting and Recording
5.4K
Reporting and recording are crucial in data documentation. The timely, thorough, and accurate documentation of facts is essential when recording patient data. Failure to record findings during an assessment or interpretation of a problem will result in loss of information and make the patient document unreliable. The reader is left with general impressions if the information is not specific. A recording is documenting data of the individual's health information in a traceable, secure, and...
5.4K
01:24Electron Carriers
91.5K
Electron carriers can be thought of as electron shuttles. These compounds can easily accept electrons (i.e., be reduced) or lose them (i.e., be oxidized). They play an essential role in energy production because cellular respiration is contingent on the flow of electrons.
Over the many stages of cellular respiration, glucose breaks down into carbon dioxide and water. Electron carriers pick up electrons lost by glucose in these reactions, temporarily storing and releasing them into the electron...
Over the many stages of cellular respiration, glucose breaks down into carbon dioxide and water. Electron carriers pick up electrons lost by glucose in these reactions, temporarily storing and releasing them into the electron...
91.5K
01:21Health Literacy
5.2K
Health literacy is an individual's or a community's capacity to comprehend, receive, read, and use relevant healthcare information and services. The World Health Organization (WHO, 2018) defines health literacy as the cognitive and social skills that determine the ability of individuals to gain access to, understand, and use information in ways that promote and maintain good health. As a result, the WHO helps individuals manage long-term health concerns, participate in preventative...
5.2K
01:14Avoidance Learning and Learned Helplessness
2.5K
Avoidance learning and learned helplessness are critical concepts in understanding behavioral responses to negative stimuli.
Avoidance learning occurs when an organism learns that a specific behavior can prevent an unpleasant outcome. For example, a student who receives a bad grade may start studying harder to avoid future poor grades. This behavior persists even when the negative outcome is no longer present. Avoidance learning is powerful because it maintains behavior in the absence of the...
Avoidance learning occurs when an organism learns that a specific behavior can prevent an unpleasant outcome. For example, a student who receives a bad grade may start studying harder to avoid future poor grades. This behavior persists even when the negative outcome is no longer present. Avoidance learning is powerful because it maintains behavior in the absence of the...
2.5K