Primary content

Software for automated quality assurance for radiotherapy patient treatment plans

Authors

L Holloway (Lead)

P Vial

G Goozee

A Xing

S Arumugam

Aims

To develop software to automate this plan checking process and validate this software in a clinical setting.

Method

In-house software, PlanChecker, was developed using the Python programing language. Inputs are: the treatment plan generated in the treatment planning system; and plan parameters within the record and verify system which is used to deliver the treatment.

To validate the tool, different types of errors were introduced for two advanced technique clinical plans by modifying the parameter assessed in the file exported from the record and verify system. The PDF report generated by PlanChecker was reviewed to verify if the program detected the errors and if these errors were recorded in the correct categories.

To validate use within the clinical setting, PlanChecker, has also has been used over the last 6 months (50 patients) in parallel with the manual process. Over 50 patient clinical plans were checked manually and with PlanChecker.

Results

In total 126 simulated errors were successfully validated to be detected by PlanChecker with the simulated errors in the clinical plans.

In comparison with manual plan check, the software is also able to check the plan parameters which are not reviewed manually. There are significant efficiency gains with the use of PlanChecker, reducing the time on average from 10 - 20 minutes for manual checking to less than 2 minutes, depending on plan complexity.

Implications

The developed software tool, PlanChecker,  is be able to correctly detect possible errors that may occur during data transfer of radiotherapy plans and can improve clinical workflow and accuracy.