MIMIC-IV
A large de-identified clinical database jointly maintained by MIT and Beth Israel Deaconess Medical Center, containing over 400,000 ICU and inpatient visit records. It is one of the most widely used public datasets in medical AI research.
Timeline (last 90 days)
The study ran a full-factorial experiment on 120 real MIMIC-IV emergency cases, covering five model families and four aggregation topologies, for a total of 20 experimental cells.
The study drew 2,048 test patient-condition pairs from TrialGPT and from question-answering datasets built on MIMIC-III and MIMIC-IV.
A new study on MIMIC-IV constructed three JEPA variants and evaluated their performance on EHR representation learning.
The study conducted systematic comparative experiments between BD and AR on the MIMIC-IV dataset.
ehr2rl is a library written in Python for converting MIMIC-IV electronic medical record data into offline reinforcement learning datasets.
ehr2rl extracts vasopressor and fluid doses from ICU input events as actions and discretizes them according to user-defined bin boundaries.
The research team constructed the MIMIC-IV-Ext-DischargeBench dataset based on the MIMIC-IV medical database, containing 477 cases covering 24 ICD disease chapters.
MIMIC-III covers data from approximately 40,000 intensive care patients during 2001–2012, while MIMIC-IV extends the time range to 2008–2019.
The research team validated HERMES on the MIMIC-III and MIMIC-IV datasets, with tasks of in-hospital mortality prediction and 30-day readmission prediction.
Compared to MIMIC-III (released in 2016), MIMIC-IV adopts a more modern data structure, divided into two main modules—hosp and icu—adds emergency department data, and extends the time coverage to 2019.
2 more timeline events
All Facts (12)
MIMIC-IV is a large-scale clinical database jointly maintained by MIT's Laboratory for Computational Physiology and Beth Israel Deaconess Medical Center, covering de-identified electronic health records of hundreds of thousands of ICU patients from 2008 to 2019.
70%UnverifiedThe study ran a full-factorial experiment on 120 real MIMIC-IV emergency cases, covering five model families and four aggregation topologies, for a total of 20 experimental cells.
50%UnverifiedThe study drew 2,048 test patient-condition pairs from TrialGPT and from question-answering datasets built on MIMIC-III and MIMIC-IV.
50%UnverifiedThe study conducted systematic comparative experiments between BD and AR on the MIMIC-IV dataset.
50%Unverifiedehr2rl is a library written in Python for converting MIMIC-IV electronic medical record data into offline reinforcement learning datasets.
50%Unverifiedehr2rl extracts vasopressor and fluid doses from ICU input events as actions and discretizes them according to user-defined bin boundaries.
50%UnverifiedThe research team constructed the MIMIC-IV-Ext-DischargeBench dataset based on the MIMIC-IV medical database, containing 477 cases covering 24 ICD disease chapters.
50%UnverifiedMIMIC-III covers data from approximately 40,000 intensive care patients during 2001–2012, while MIMIC-IV extends the time range to 2008–2019.
50%UnverifiedThe research team validated HERMES on the MIMIC-III and MIMIC-IV datasets, with tasks of in-hospital mortality prediction and 30-day readmission prediction.
50%UnverifiedCompared to MIMIC-III (released in 2016), MIMIC-IV adopts a more modern data structure, divided into two main modules—hosp and icu—adds emergency department data, and extends the time coverage to 2019.
50%UnverifiedThe dataset used in this study contains 6,882 patients and 54,694 time windows, with the experimental setup using 24 hours of historical data to predict physiological indicator changes over the next 6 hours.
50%UnverifiedA new study on MIMIC-IV constructed three JEPA variants and evaluated their performance on EHR representation learning.
50%