Click here to

Session: Multi-Disciplinary General ePoster Viewing [Return to Session]

A High-Throughput, High-Availability Automated Planning Tool for Radiotherapy Clinics in Low-Resource Settings

B Marquez1,2*, S Gay2, R Douglas2, L Zhang2, K Huang1,2, M El Basha1,2, C Cardenas1,2, L Court1,2, (1) University of Texas MD Anderson UTHealth Graduate School of Biomedical Sciences, (2) University of Texas MD Anderson Cancer Center, Houston, TX


PO-GePV-M-9 (Sunday, 7/25/2021)   [Eastern Time (GMT-4)]

Purpose: To assess the throughput and reliability of a web-based, automated radiotherapy planning tool (Radiation Planning Assistant, RPA) designed to support low-resource clinics.

Methods: The RPA architecture consists of several, multi-capacity computing modules that process patients in a serial manner. Baseline timings for each module were measured by processing 25 cervix and 25 head & neck (H&N) patient datasets through the entire workflow. Each module was modeled in a manufacturing discrete event simulator (ManPy), and programmed with their respective timing data, to evaluate serial and parallel workflows. Model accuracy was evaluated by comparing the simulator’s completion times for single- and multi-patient queues to those of the real RPA system. Finally, many module downtime scenarios were simulated to determine their impact on baseline performance of the RPA's daily throughput.

Results: An independent t-test indicated that mean processing times were not significantly different between the RPA and the model (p>0.05). The model gave an expectation value for plans successfully processed in 24 hours: 483 cervix plans, 255 H&N contours, or 258 H&N plans when all systems are operating. Cervix plan generation remained within 5% of its baseline throughput until any given module (except the plan/dose QA module) went down for 3+ hours. H&N contour generation remained within 5% of its baseline when either of its two contouring module's downtime did not exceed 1 hour. H&N plan generation remained within 5% of its baseline until at least 2 of 5 available VMAT optimization modules’ downtime exceeded 1 hour, or any other module downtime exceeded 3 hours. Plan calculation and report generation module downtimes had <5% effect on output through 4 hours downtime.

Conclusion: The RPA architecture is robust to downtime of its individual modules, thus having the ability to reliably address resource shortage in radiotherapy clinics.



    Not Applicable / None Entered.


    IM/TH- Formal Quality Management Tools: General (most aspects)

    Contact Email