K-MIMIC
Korea Medical Information Mart for Intensive Care
Introduction
We are pleased to introduce our K-MIMIC database. K-MIMIC (Korea Medical Information Mart for Intensive Care) is an extensive health-related database derived from patients who were admitted to the intensive care units (ICUs) of hospitals in Korea. This database has been established through the support of a grant from the Korea Health Technology R&D Project, managed by the Korea Health Industry Development Institute (KHIDI), and funded by the Ministry of Health & Welfare, Republic of Korea (grant number: HI21C1074).
K-MIMIC comprises a wide range of data, including electronic medical records, medical images, and vital sign information, serving as a crucial resource for research and development in the field of intensive care medicine.
"K-MIMIC Dataset (Customized Version) for Korea Clinical Datathon 2024"
Korea Clinical Datathon 2024 is an event closely associated with the K-MIMIC project, which aims to collect a total of 1 million data entries. For the purpose of this Datathon, we are pleased to provide you with a selected subset of this extensive dataset.
Summary of Data Collection Period and Record Count
Electronic Medical Records (EMR)
The K-MIMIC database's electronic medical records (EMR) encompass detailed medical information for patients admitted to intensive care units (ICUs). While an extensive dataset was collected from multiple centers, only a curated subset of this comprehensive database is being made available for the Korea Clinical Datathon 2024. The Datathon dataset will include data covering a period of three months. The collection and processing of this data were carried out using the IMPACT Platform.
The EMR data is categorized into five sections:
Core, Hosp, ICU, ED, and HIS, with each category comprising the following tables:
Core : ADMISSIONS, PATIENTS, TRANSFERS
Hosp : DIAGNOSES_ICD, D_LABITEMS, D_TESTITEMS, EMAR, EMAR_DETAIL, LABEVENTS, MICROBIOLOGYEVENTS, PRESCRIPTIONS,
PROCEDURES_ICD, SERVICES, TESTEVENTS
ICU : CHARTEVENTS, DATETIMEEVENTS, D_ITEMS, ICUSTAYS, INPUTEVENTS, OUTPUTEVENTS, PROCEDUREEVENTS
ED : EDSTAY, MEDRECON, TRIAGE
HIS : HOSPITAL
which includes data from 3,738 ICU stays of 3,234 patients across Multi-Center.
Patients (unit: person)
ICU Stays (unit: case)
Medical Images
Modality : X-ray
For the Korea Clinical Datathon 2024, only a customized subset of this extensive database will be made available. This database encompasses a vast amount of medical data collected from a single hospital. The Datathon dataset will include data covering a period of three months and the collection and processing of this data were conducted using the IMPACT Platform. This dataset features medical images, including X-rays, with a total of 452 cases.
Required Fields for Finding Medical Imaging Data that Matches from EMRs
We extract medical imaging data belonging to a specific modality and category using EMR. The required modality is Medical Imaging (X-ray), the required category is Abdomen Supine, and the required EMR tables are D_TESTITEMS (Hosp), TESTEVENTS (Hosp).
D_TESTITEMS : This pertains to information about imaging tests and intra-departmental functional tests.
TESTEVENTS : This pertains to information about Patient Imaging Test Records.
To search for the modality and category of the hospital in the D_TESTITEMS table, we refer to the following columns:
ITEMID: This column stores the test code, generated for data tracking.
LABEL: This column contains the test name.
CATEGORY: This column includes information about the test category.
EDI_CODE: This column holds the EDI code.
To search for the modality and category of the hospital in the TESTEVENTS table, we refer to the following columns:
SUBJECT_ID: Patient identifier for patients with a history of ICU stay.
CHARTTIME: Temporal information about the patient related to the test results (e.g., imaging date, admission date).
FILE_PATH: The DICOM file path of the respective imaging test results within the central platform.
Bio-signals
Bio-signals(.vital) collected using the Vital Recorder installed on ICU patient monitors.
The Bio-signal dataset is a comprehensive collection of bio-signals (vital signs) obtained from ICU patients with a total of 112 cases. These data were collected using the Vital Recorder installed on ICU patient monitors at a single hospital. Although an extensive database was collected, only a customized subset of this data will be made available for the Datathon. This dataset serves as a valuable resource for researchers and healthcare professionals aiming to study and analyze bio-signals from ICU patients. The comprehensive data collection facilitates advancements in medical research and the improvement of patient care.
Required Fields for Finding Bio-signal Data that Matches from EMRs
We extract Bio-signal data belonging to a specific modality and category using EMR. The required EMR tables is ICUSTAYS.
ICUSTAYS : ICU Admission and Discharge Records. The information pertains to a patient's ICU admissions and discharges, with each stay from admission to discharge being represented by a single ID.
To search for the modality and category of the hospital in the ICUSTAYS table, we refer to the following columns:
SUBJECT_ID: Patient Identifier for Patients with a History of ICU Stay.
STAY_ID: ICU Admission-Discharge ID: A unique key to distinguish ICU admission and discharge records.
INTIME: Date and Time of ICU Admission.
OUTTIME: Date and Time of ICU Discharge.