Click here to

Session: Professional General ePoster Viewing [Return to Session]

An Implementation Strategy for Introducing Automated Treatment Planning Into the Clinic

J Jackson*, H Kang, J Wick, J Roeske Loyola University Medical Center, Maywood, IL

Presentations

PO-GePV-P-50 (Sunday, 7/25/2021)   [Eastern Time (GMT-4)]

Purpose: The purpose of this study is to design an implementation strategy for introducing new automated treatment planning scripts into clinical use. Deployment of automated planning for clinical use requires thoughtful preparation that ensures safety and consideration of additional failure modes introduced by automation of treatment planning tasks.

Methods: The Eclipse Scripting Application Programming Interface (ESAPI) was used to develop an automated treatment planning script to be used for radiotherapy patients. A workflow was mapped to assist in clinical implementation of this script. The steps in this workflow involved: assessing clinical need for automation, discussions with dosimetry staff regarding preservation of current planning workflow, iterative testing for bugs and performance, code review by another physicist or programmer, introduction of unexpected inputs to find potential hidden failure modes, and training for physics and dosimetry staff. After release for clinical use, we also considered feedback from dosimetry staff for new features and updates as well as conducted ongoing physics review to assess the need for script changes due to adjustments in treatment planning strategy.

Results: By following a regimented implementation procedure, we were able to deploy an automated planning script which is customized to our site’s particular planning workflow. This procedure helped discover several failure modes that may have been missed otherwise. Blocks of code were added to prevent these failure modes which included potential incorrect isocenter placement and transient bugs that could have caused program crashes with rare, unexpected inputs.

Conclusion: Automated treatment planning has the ability to reduce time spent planning, standardize plan quality, and reduce the probability of failure modes that are more likely to occur with manual planning. The time spent evaluating a new automated planning script is considerable, and the steps involved should be carefully planned prior to implementation.

ePosters

    Keywords

    Treatment Planning, ESAPI, Automation

    Taxonomy

    TH- External Beam- Photons: Treatment planning using machine learning/Knowledge Based Planning/automation

    Contact Email

    Share: