Program Information
Cloud Based Monitoring of X-Ray Angiography Equipment Performance Based On Automated Daily Sampling of Key Performance Indicators
J Sjoeberg1*, A Omar1 , R Bujila1 , P Nowik1 , I Olafsson2 , G Poludniowski1 , (1) Karolinska University Hospital, Stockholm, (2) Image Owl Inc., Reykjavik
Presentations
SU-G-IeP3-15 (Sunday, July 31, 2016) 5:00 PM - 5:30 PM Room: ePoster Theater
Purpose: To investigate the feasibility of deploying a novel method for the daily monitoring of x-ray Angiography (XA) equipment performance as a cloud-based service.
Methods: From previous work, we have developed a novel image based method of testing the performance of XA equipment on a daily basis. A pre-programmed protocol on an XA system is used and images are automatically analyzed using software developed in-house, resulting in a set of Key Performance Indicators (KPI’s) that can be tracked over time. The analysis software was deployed in the cloud and a routing rule was set up to automatically transfer images from the XA systems to the cloud. A web based user interface was developed with access to the KPI’s for studying auto generated reports, trends and anomalies. Technologists were instructed on how to perform the daily scans with a remotely hosted instructional video and have actively participated in this endeavor for the last 7 months.
Results: Quantitative performance tests have been automated including data aggregation, management, analysis and visualization which enables sampling of KPI’s on a daily basis and the developed solution has been migrated to the cloud. The daily performance check routine takes ~6 minutes per XA system in total including the transfer of images, analysis and reporting. In one instance a significant artifact and a defective pixel has been identified. A generally stable system performance has otherwise been observed as indicated by the process control charts.
Conclusion: A cloud based solution for monitoring the performance of XA equipment has been developed. The solution facilitates efficient data collection, enabling useful trend analysis on large data sets.
Funding Support, Disclosures, and Conflict of Interest: This project was funded by Vinnova (Swedish Innovation Agency), grant number 2012-03693, and was carried out in collaboration with Image Owl, Inc. (Reykjavik, Iceland).
Contact Email: