Radiology Workflow Simulation Dataset: Synthetic Patient Flows, Event Logs, Exam Records, and Resource Utilization Metrics based on a Published Radiology Workflow Model
Description
This dataset contains a complete synthetic reproduction of radiology patient flows generated through a custom discrete-event simulation model of a medium-sized hospital. The simulation was performed using CrogMagnon [1], a simulation engine developed within the NEAT-AMBIENCE Project [2], led by the University of Zaragoza (Spain), and with potential improvements and extensions in a follow-up project at that university. The simulated workflow is based on a real radiology patient-flow description published in a peer-reviewed scientific article analyzing operational performance in a Belgian hospital radiology department [3]. The original study outlines the sequence of administrative and imaging steps, bottlenecks, modality-specific processes, and typical resource constraints. These elements have been faithfully reproduced and parameterized to create a realistic synthetic representation aligned with real-world radiology practice. The dataset includes three core CSV files describing patient-level, exam-level, and event-level trajectories, as well as detailed resource-usage logs for each imaging device and structured scenario metadata files documenting all simulation parameters and assumptions. The simulation covers a 7-day period beginning on January 1st, 2023. Dataset structure ------------------- The dataset contains three main CSV files and three supporting folders: • cases.csv – one row per simulated patient case. • exams.csv – one row per radiology exam. • events.csv – one row per event in each patient’s workflow. Additionally: • resource_usage/ – time-series logs of resource utilization (every 5 minutes). • metadata/ – scenario metadata files (arrival patterns, patient types, resource configuration). • images/ - containing simulated-radiology-process.png, an image describing the simulated process. A more detailed description of each file is provided in the accompanying document “dataset_details.txt”. Intended Use --------------- This dataset is suitable for research and investigation in healthcare operations, radiology workflow analysis, and discrete-event simulation. But it can also be used for machine learning experiments, such as prediction of waiting times and length of stay, analysis of workflow trajectories, benchmarking algorithms for workflow optimization, etc. A more detailed overview of the usescases from this dataset can be seen in the attached file: 'intended_uses.txt' References ------------- [1] https://webdiis.unizar.es/~silarri/prot/Cro-Magnon [2] https://webdiis.unizar.es/~silarri/NEAT-AMBIENCE [3] van Hulzen, G., Martin, N., Depaire, B., & Souverijns, G. (2022). Supporting capacity management decisions in healthcare using data-driven process simulation. Journal of Biomedical Informatics, 129, 104060. https://doi.org/10.1016/j.jbi.2022.104060
Files
Institutions
- Universidad de Zaragoza Escuela de Ingenieria y Arquitectura
Categories
Funders
- Departamento de Ciencia, Universidad y Sociedad del Conocimiento del Gobierno de Aragón (Government of Aragon)
- MICIU/AEI/10.13039/501100011033 (Spanish State Research Agency)Grant ID: PID2020-113037RB-I00
- Group Reference T64_23R, COSMOS research group