Purpose or Case Report: MRI scans are long, and scan durations are unpredictable. Combined with challenges of inter-departmental coordination, this leads to poorer operational efficiency, increased need for sedation (especially in a pediatric population), increased wait time, and overall poorer patient care. There is a need for active, automatic tracking of scan log in a MRI machine to 1) determine operational efficiency, 2)disease specific scan information, 3) identifying patient specific scan metrics.
Currently, there exist no customizable open-source solution that can automatically obtain information from MRI scanners. The purpose of this study is to demonstrate implementation of a novel, customizable QI tool that can automatically extract scan log data from a MRI scanner in a pediatric setting.
Methods & Materials: ‘LogServer’ (https://yarra.rocks/doc/server/yarralogserver/) was developed for use in Siemens MRI machines. This uses the ‘Yarra’ framework – an open source framework for complex MRI reconstructions. ‘LogServer’ monitors sequences, parses MRI sequences into discrete exams, generates ‘EPIC-formatted’ patient names, monitors scanner health, and tags exams with body region information. All extracted data were stored in ‘PostgreSQL’ - an open source database. ‘ReDASH’, an open source dashboard, was used to create customizable dashboards. In this study, for a clinical MRI scanner, we looked at 1) Daily scanner efficiency, defined as ratio of total hours scanner unused time to total hours (8am-7pm workday), and 2) Scan efficiency, defined as time when scanner was running to total time when the patient was on table.
Results: ‘LogServer’ was successfully able to automatically extract scan logs from MRI machines on an hourly basis, extract all relevant information, and visualize them in needed format using a remote PHI secure web browser. Figure 1 (https://drive.google.com/open?id=1fvyrSjq96EcHaAv9GenX1v4b070tFP3l) demonstrates the scanner schedule for any given day. Figure 2 (https://drive.google.com/open?id=1O7fxAIhWrAeqFVr3qRW5QyMGaN3F0mnX) shows the scan metrics for a selected patient, including total scan time, idle time during a scan session, as well time for adjustments in scanner. Our daily scanner efficiency for a week was 64% while scan efficiency was 76.3 ± 10%.
Conclusions: We demonstrated a robust customizable QI tool that lets continuous automatic monitoring of MRI scanner. This is a useful tool for hospital administrators, clinical managers, clinical leaders in their informed decision making.