Dark Mode
Infrared Thermography (IRT) Clinical Accuracy Database
Data Science and Analytics
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
includes 124 variables across six files. These variables detail temperature measurements from specific facial locations, along with crucial demographic and environmental metadata.
Key variables include:
- SubjectID: A unique, de-identified code for each volunteer.
- T_offset1 to T_offset4: Temperature offsets recorded across four measurement rounds.
- Facial Temperature Variables: 26 distinct measurements (e.g., T_CRmax, T_CLmax for maximum temperature at the right/left inner canthus, T_Max for overall maximum facial temperature) recorded for four rounds.
- Oral Temperature Variables:
aveOralF(average oral temperature, fast mode) andaveOralM(average oral temperature, monitor mode). - Demographics: Gender, Age (provided in groups, e.g., 21-30, 60 and above), and Ethnicity.
- Environmental/Operational Parameters:
T_atm(ambient temperature), Humidity, Distance, Cosmetics (whether applied), Time, and Date.
Distribution
The data is organised into six separate CSV files, based on the camera used (FLIR or ICI) and the ambient temperature range during collection (Group 1: 20.0 to 24.0 °C; Group 2: 24.0 to 29.0 °C).
Data for 1020 subjects is available for the FLIR thermograph (IRT-1), and data for 1009 subjects is available for the ICI thermograph (IRT-2). Approximately 11% of the subjects included in the analysis exhibited a reference temperature above 37.5 °C. For instance, the
FLIR_group2.csv file contains 478 records. The data files are in CSV format.Usage
This database is valuable for researchers seeking to advance thermal imaging systems. Ideal applications include:
- Correlation Analysis: Determining the relationship between temperature values taken at different facial sites and measured by different devices.
- Accuracy Evaluation: Establishing and validating metrics for IRT clinical accuracy, especially in high-throughput fever screening scenarios.
- Parameter Impact Study: Assessing the influence of factors like ambient temperature, humidity, gender, and age on the accuracy of thermal measurements.
- Algorithm Development: Training and testing algorithms for the automated detection of elevated body temperature based on facial thermal patterns.
Coverage
The data was collected during a clinical study performed in a reasonably well-controlled environment, although seasonal variations were present.
- Technology: Two distinct Infrared Thermographs (FLIR and ICI) were used.
- Demographic Scope: While demographic information (Gender, Age, Ethnicity) is recorded, the age range is notably limited; 95% of subjects were under 30 years of age, with roughly half being 18–21 years old. Age information is grouped to comply with HIPAA regulations.
- Time Range: Measurement dates are included, which supports analysis of seasonal effects on thermography readings.
License
Creative Commons Zero 1.0 Universal Public Domain Dedication
Who Can Use It
- Medical Device Developers: To benchmark the performance and accuracy of new thermal imaging technology against standardised clinical data.
- Academics in Biomedical Engineering: For thesis work involving signal processing, image registration, and heat transfer modelling related to the human face.
- Governmental/Regulatory Bodies: To inform standards and guidelines related to the deployment and implementation of IRTs for public health screening.
Dataset Name Suggestions
- Clinical Facial and Oral Temperature Data for Elevated Body Temperature Screening
- Infrared Thermography (IRT) Clinical Accuracy Database
- De-identified Human Thermal and Oral Temperature Measurements
Attributes
Original Data Source: Infrared Thermography (IRT) Clinical Accuracy Database
QUALITYLoading...