Machine learning-based predictions of gamma passing rates for virtual specific-plan verification based on modulation maps, monitor unit profiles, and composite dose images

QUINTERO, Paulo, BENOIT, David, CHENG, Yongqiang, MOORE, Craig and BEAVIS, Andrew (2022). Machine learning-based predictions of gamma passing rates for virtual specific-plan verification based on modulation maps, monitor unit profiles, and composite dose images. Physics in Medicine & Biology, 67 (24): 245001.

[img]
Preview
PDF
pdf.pdf - Published Version
Creative Commons Attribution.

Download (1MB) | Preview
Official URL: https://iopscience.iop.org/article/10.1088/1361-65...
Open Access URL: https://iopscience.iop.org/article/10.1088/1361-65... (Published version)
Link to published version:: https://doi.org/10.1088/1361-6560/aca38a

Abstract

Machine learning (ML) methods have been implemented in radiotherapy to aid virtual specific-plan verification protocols, predicting gamma passing rates (GPR) based on calculated modulation complexity metrics because of their direct relation to dose deliverability. Nevertheless, these metrics might not comprehensively represent the modulation complexity, and automatically extracted features from alternative predictors associated with modulation complexity are needed. For this reason, three convolutional neural networks (CNN) based models were trained to predict GPR values (regression and classification), using respectively three predictors: (1) the modulation maps (MM) from the multi-leaf collimator, (2) the relative monitor units per control point profile (MUcp), and (3) the composite dose image (CDI) used for portal dosimetry, from 1024 anonymized prostate plans. The models’ performance was assessed for classification and regression by the area under the receiver operator characteristic curve (AUC_ROC) and Spearman’s correlation coefficient (r). Finally, four hybrid models were designed using all possible combinations of the three predictors. The prediction performance for the CNN-models using single predictors (MM, MUcp, and CDI) were AUC_ROC = 0.84 ± 0.03, 0.77 ± 0.07, 0.75 ± 0.04, and r = 0.6, 0.5, 0.7. Contrastingly, the hybrid models (MM + MUcp, MM + CDI, MUcp+CDI, MM + MUcp+CDI) performance were AUC_ROC = 0.94 ± 0.03, 0.85 ± 0.06, 0.89 ± 0.06, 0.91 ± 0.03, and r = 0.7, 0.5, 0.6, 0.7. The MP, MUcp, and CDI are suitable predictors for dose deliverability models implementing ML methods. Additionally, hybrid models are susceptible to improving their prediction performance, including two or more input predictors.

Item Type: Article
Additional Information: ** From IOP Publishing via Jisc Publications Router ** Licence for this article: http://creativecommons.org/licenses/by/4.0/ **Journal IDs: eissn 1361-6560 **Article IDs: publisher-id: pmbaca38a; manuscript: aca38a; other: pmb-114101.r1 **History: published 21-12-2022; open-access 06-12-2022; published_online 06-12-2022; oa-requested 16-11-2022; accepted 16-11-2022; rev-recd 31-10-2022; submitted 24-09-2022
Uncontrolled Keywords: Paper, machine-learning, radiotherapy, CNN, gamma-passing-rates
Identification Number: https://doi.org/10.1088/1361-6560/aca38a
SWORD Depositor: Colin Knott
Depositing User: Colin Knott
Date Deposited: 07 Dec 2022 13:20
Last Modified: 12 Oct 2023 08:33
URI: https://shura.shu.ac.uk/id/eprint/31122

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year

View more statistics