UTA7: Radiologist-Curated DICOM Medical Imaging for HCI Research

UTA7 consists of DICOM medical imaging files provided by radiologists for research published in the International Journal of Human-Computer Studies. Developed by MIMBCD-UI and last updated in June 2022, the dataset focuses on the intersection of clinician interaction and medical imaging technology.

Use Cases

• Developing user-centered design interfaces for physicians by testing DICOM file rendering and interaction
• Training deep learning models for medical image classification using radiologist-verified samples
• Conducting human-computer interaction (HCI) experiments to measure clinician efficiency when navigating DICOM metadata

Strengths

• Expert-sourced data provided directly by professional radiologists
• Peer-reviewed provenance via the International Journal of Human-Computer Studies
• Standardized DICOM format ensures compatibility with clinical PACS and viewing software

Limitations

• Unknown sample size and demographic distribution
• Lack of explicit metadata documentation or data dictionaries in the public repository
• Potential selection bias based on the specific clinician workflows studied in the original research

Provenance

Source

MIMBCD-UI / International Journal of Human-Computer Studies

Collection Method

Annotated and provided by radiologists for research purposes

Freshness

The dataset was last updated in June 2022 and is tied to a static academic publication.