Automation of Dosimetric Data Extraction for Gynecologic Intracavitary Brachytherapy Plans using the Eclipse Scripting Application Programming Interface

Purpose: Radiation oncology treatment plans are complex and require many dose metrics that are useful in both clinical plan quality judgement and physics research. Data extraction from treatment plans can be a cumbersome, time consuming, and error-prone process. The Eclipse Scripting Application Programming Interface (ESAPI) Microsoft .NET framework and its ability to access the ARIA (a recording and verifying system developed by Varian) treatment planning database makes it a powerful tool for clinicians and researchers alike to incorporate automation into data collection. Automation of data collection allows for more efficient and error-free collection of dosimetric data while eliminating a simple and cumbersome bottleneck process in clinical research. The main objective of our study was to develop an ESAPI script to automate data collection and demonstrate the feasibility of automated data collection for high dose rate (HDR) gynecologic intracavitary brachytherapy plans.

Methodology: Through proper scripting work, ESAPI allows to read and access the treatment plan data in Eclipse, thereby enabling the retrieval of information such as plan, structure, dose, and dose-volume histogram (DVH) data from the ARIA database. By utilizing Eclipse’s capability to support C# programming to develop ESAPI scripts, an ESAPI script capable of receiving a list of patients and creating CSV files with relevant extracted treatment plan data was developed. We then applied the ESAPI script to a database of cervical cancer patients who underwent HDR tandem and ovoid brachytherapy as part of their treatment. Eclipse commands were accessed to create a CSV file with data points along the DVH and a CSV dataset consisting of D95, D90, D50, D10, D5, V100, V95, V90 of both the high-risk clinical target volumes (HR-CTV) and intermediate-risk clinical target volumes (IR-CTV).

Results: The developed ESAPI script automatically generates a CSV file of requested dosimetric data and a CSV file of data points along the DVH for the structure of interest. The desired dose metrics of both the HR-CTV and IR-CTV were extracted. The average time required to generate data with manual review was 10 minutes per patient, while automatically, it took less than 5 seconds per patient, which is a 99% reduction in time relative to manual review.

Conclusions: This study highlights the usefulness of a newly developed ESAPI script that allows for automated data collection from complex HDR brachytherapy plans. Automation has simplified the data extraction process and provides an efficient and accurate method of extracting dosimetric data while greatly decreasing time spent in chart review and reducing possible human errors.

References: