MIMIC-III
A publicly available intensive care medical information dataset containing a large volume of de-identified patient clinical records, widely used in medical AI research.
Timeline (last 90 days)
The study drew 2,048 test patient-condition pairs from TrialGPT and from question-answering datasets built on MIMIC-III and MIMIC-IV.
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.
MIMIC-III covers data from approximately 40,000 intensive care patients during 2001–2012, while MIMIC-IV extends the time range to 2008–2019.
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.
MIMIC-III is a large de-identified electronic health record database maintained by MIT's Laboratory for Computational Physiology, containing clinical data for approximately 40,000 patients in the intensive care unit of Beth Israel Deaconess Medical Center between 2001 and 2012.
ClinicalBERT is further trained on electronic health record data such as MIMIC-III, mastering clinical abbreviations and non-standard writing patterns
All Facts (6)
The study drew 2,048 test patient-condition pairs from TrialGPT and from question-answering datasets built on MIMIC-III and MIMIC-IV.
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%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%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%UnverifiedMIMIC-III is a large de-identified electronic health record database maintained by MIT's Laboratory for Computational Physiology, containing clinical data for approximately 40,000 patients in the intensive care unit of Beth Israel Deaconess Medical Center between 2001 and 2012.
50%UnverifiedClinicalBERT is further trained on electronic health record data such as MIMIC-III, mastering clinical abbreviations and non-standard writing patterns
50%