Program Information
Platform for MRI Quality Control, Automated Image Analysis, and Monitoring
J M Wilson*, S Robertson, E Samei, Duke University Medical Center, Durham, NC
Presentations
TU-D-601-4 (Tuesday, August 1, 2017) 11:00 AM - 12:15 PM Room: 601
Purpose: Accurately and continuously assessing image quality is foundational in any radiology quality control program. Daily and weekly MRI quality control (QC) image acquisition and analysis is typically delegated to MRI technologists. However, for a multisite hospital with numerous technologists who rotate through multiple clinics and cover three shifts, ensuring consistent QC requires oversight. Thus, a centralized platform for analysis of QC images data across scanners and longitudinally can ensure consistent and reliable data. The goal of this work was to implement a centralized platform for MRI QC.
Methods: Phantoms were scanned by technologists daily and weekly on eighteen different systems consisting of two vendors, thirteen 1.5 T and five 3.0 T, across three hospitals. Images were sent to a QC server where a validated algorithm automatically measured standard metrics: table positioning, geometry, uniformity, SNR, ghosting, slice positioning, spatial resolution, and low-contrast resolution. QC results were databased along with annual evaluation results and system-specific tolerances. Visual checklists were submitted through a webform. Alerts were emailed to medical physicists if a value exceeded its tolerance, and technologists were reminded if a system’s QC was overdue. An internal webpage shows QC statuses, metrics with trend plots, and visualizations of analyzed data.
Results: Since the platform’s inception in January 2014, over 2,235 weekly and 1,023 daily (only interventional) studies have been received, analyzed, and databased along with respective visual checklists. The oldest systems each have 150-155 studies, and the newest systems have 30-35 studies. Weekly QC compliance is currently >96%, and daily QC compliance is >99%.
Conclusion: Our centralized MRI QC platform automates image quality analysis and simplifies reporting and monitoring. This platform can help identify compliance issues, automatically communicates important results to physicists, integrates the results of annual physics evaluations with QC monitoring, and facilitates the deployment of new QC metrics.
Contact Email: