Brain Imaging Dataset
This dataset comprises electronic health records (EHR) of patients accompanied by brain CT and MRI images. The patients are evenly divided into two groups: Case and Control. Case indicates that there is a confirmed brain cancer diagnosis of this patient, while Control means no diagnosis of cancer. The dataset includes approximately 3,000 patients, with approximately 1,500 having CT images and 1,500 having MRI images. In each image type, patients are further split into 750 Cases and 750 Controls. The structured EHR data encompasses a wide range of information, including demographics, encounters, diagnoses, lab results, medications, vital signs, problem lists, procedures, and social determinants of health events. Additionally, radiology and pathology reports are available.
Who are the Patients?
Breakdown of patients demographic, emphasis on health disparities and equity.
Brain CT
Brain MRI
| Domain | Example Variables | Description |
|---|---|---|
|
Person |
Person ID, Sex, Race, Ethnicity, Age_Group, 3 digit Zip Code, Deceased Date Time |
Demographic characteristics from EHR, along with mortality information. |
|
Problem |
Person ID, Standard_Code, Problem_Status, Onset Date Time, Update Date Time |
Problem documented by providers. Most are SNOMED Code. Update time shows the last update time for this problem, and status shows whether this problem is active or has been resolved. |
|
Diagnosis |
Person ID, Encounter ID, Diagnosis Type, Diagnosis Date Time, Diagnosis Code |
All Medical Diagnosis Records(History and Follow-up) for all patients who are in this cohort at encounter level. Diagnosis code could be ICD9 or ICD10. |
|
Encounter |
Person ID, Encounter ID, Registration Date Time, Discharge Date Time, Reason for Visit, Encounter Type, Height, Weight, BMI, Tobacco Use, Discharge Disposition, Died at discharge Indicator |
Information about healthcare utilization (e.g., ER, inpatient, and outpatient visits), along with the most recent Height/Weight/BMI/Tobacco event results and discharge disposition for that visit. |
|
Medication |
Person ID, Encounter ID, Order Mnemonic, Order Status, Original Order Time, Projected Stop Time, Order Flag, Order Detail, Catalog Name |
Lists of all medications with the catalog name, order status, order/projected stop date time and order details. Catalog uses Cerner Multum Drug Information. |
|
Procedure |
Person ID, Procedure Code, Encounter ID, Procedure Description, Service Date Time |
Records procedures performed on patients, with procedure codes, descriptions, and service dates on Encounter level, which are important for tracking patient treatments. |
|
Labs |
Person ID, Encounter ID, Lab Name, Result Unit, Result Time, Result Value, Range |
Contains lab test results, with details such as lab name, result units, times, values, and normal ranges, essential for monitoring patient health. |
|
Vitals |
Person ID, Encounter ID, Vital Name, Result Unit, Result Time, Result Value, Range |
Includes vital sign measurements, with details on the type of vital signs, units, times, values, and normal ranges, crucial for assessing patient status. |
|
Social Info |
Person ID, Category, Description, Response |
Social history categories include tobacco, alcohol, exercise, health/nutrition, employment/education, Substance Use, and so on. |