Related Experiment Video
Updated: Jun 28, 2026
Methods to Test Visual Attention Online
Published on: February 19, 2015
Human Observer Net: A Platform Tool for Human Observer Studies of Image Data
1From the Department of Radiology, Charité-Universitätsmedizin Berlin, corporate member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Charitéplatz 1, 10117 Berlin, Germany (U.G., P.J.); Data Analytics and Computational Statistics, Hasso Plattner Institute, Digital Engineering Faculty, University of Potsdam, Potsdam, Germany (U.G.); and Berlin Institute of Health, Berlin, Germany (P.J.).
A new open-source software platform enhances medical imaging observer studies. This user-friendly tool offers flexible study designs, platform independence, and multicenter capabilities for efficient research.
Area of Science:
- Medical Imaging
- Human Observer Studies
- Software Development
Background:
- Existing software for human observer studies in medical imaging lacks flexibility in study design, platform independence, multicenter use, and assessment methods.
- Current applications are often not open-source, limiting accessibility and expandability for researchers.
Purpose of the Study:
- To develop a user-friendly, open-source software platform for efficient human observer studies in medical imaging.
- To provide flexibility in study design, platform independence, and multicenter capabilities.
Main Methods:
- Developed an open-source web application for human observer imaging studies.
- Implemented interfaces for study creation, participation, and results management.
- Evaluated software usability with 14 radiologists using the System Usability Scale and tracked response times.
Main Results:
- The software supports various analysis methods including visual grading analysis (VGA), multiple-alternative forced-choice (m-AFC), and receiver operating characteristic (ROC) analyses.
- Mean reader response times ranged from 5.8 to 8.7 seconds per image/set across different study designs.
- Achieved a high mean System Usability Scale score of 83 ± 11, indicating excellent usability.
Conclusions:
- A user-friendly, efficient, open-source application for human reader experiments has been successfully developed.
- The platform offers versatility in study design, platform independence, and multicenter usability for medical imaging research.
