Open Collections

Borden, Bailey

University of British Columbia

Purpose: Artificial intelligence (AI) autocontouring tools are becoming readily available, and can also be created using published deep neural network (DNN) frameworks. With the labour demand required to peer review cases, contouring peer review may not occur before the beginning of treatment. This work investigates the use of AI for autocontouring to aid within the breast radiotherapy contouring peer review process. The use of AI within contour peer review may aid with labour demands and allow patient cases to be reviewed earlier in treatment. Methods: Using open-source deep learning libraries, DNNs were trained to segment the heart, the left lung, and the right lung. The model architectures explored in this work are U-Net and LinkNet. The trained architectures were compared to the commercial algorithm Limbus AI. The trained U-Net model architecture was selected to demonstrate uncertainty quantification using conformal prediction, while Limbus AI was used to show an approach to model uncertainty using test time augmentation (TTA). Results: For multiclass segmentation explored, U-net and Linknet achieved a mean intersection over union (IoU) of 0.924 and 0.910 respectively. When the trained algorithms were compared to the commercial algorithm, U-Net and LinkNet achieved mean IoUs of 0.914 and 0.911, while Limbus AI achieved a mean IoU of 0.883. Using the conformal prediction methodology on U-Net, uncertain pixels within the model's predictions were found to be on the boundaries of structures. When testing the uncertainty of the Limbus AI using TTA, a standard error of 4.95% was observed on the heart predictions, 3.54% on the left lung predictions, and 3.35% on the right lung predictions. Conclusion: Based on the literature, evaluating contours using both area and distance based metrics is recommended for peer review quality assurance. Testing of a selected model using a sample of data representing the potential patient demographic is recommended. Additionally, testing the uncertainty of a selected model using unseen data is recommended, to provide model insight. The appropriate use of these AI algorithms could aid in the contouring peer review process.

Text

2024-12-05

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/89784

Medical Physics

University of British Columbia

UBCO

http://creativecommons.org/licenses/by-nc-nd/4.0/