ESTRO 2024 - Abstract Book

S2919

Interdiscplinary - Other

ESTRO 2024

contributing factor for cases flagged for manual review, but yet considered acceptable following review. With only one true violation identified in the datasets reviewed, further assessment of the tool in practice is needed to determine the reliability of the CQA tool in detecting true negatives.

Auto QA Flagged for Review

1

21

Auto QA Acceptable

0

12

RO QA Violation

RO QA Acceptable

Figure 2: Confusion matrix indicating the number of cases the automated CQA tool flagged for review compared to number the RO deemed acceptable during manual review.

Conclusion:

We have demonstrated practical feasibility of implementing an automated CQA tool within the active NINJA trial. In this setting the tool achieved high specificity and demonstrated the ability to avoid a manual review in 35% of patients. Future work will assess the tool on additional patients recruited to the trial and consider the most appropriate approach for adapting the criteria for when a manual review is needed.

Keywords: contour quality assurance, implementation

References:

[1] J. Martin et al., “TROG 18.01 phase III randomised clinical trial of the Novel Integration of New prostate radiation schedules with adJuvant Androgen deprivation: NINJA study protocol,” BMJ Open, vol. 9, no. 8, pp. 1–10, 2019, doi: 10.1136/bmjopen-2019-030731. [2] H. Min et al., “Clinical target volume delineation quality assurance for MRI-guided prostate radiotherapy using deep learning with uncertainty estimation,” Radiotherapy and Oncology, vol. 186, p. 109794, 2023, doi: 10.1016/j.radonc.2023.109794.

1593

Mini-Oral

Can satellite centres help reduce the carbon footprint of external beam radiotherapy?

Robert W Chuter