DoseGAN: a generative adversarial network for synthetic dose prediction using attention-gated discrimination and generation

Abstract Deep learning algorithms have recently been developed that utilize patient anatomy and raw imaging information to predict radiation dose, as a means to increase treatment planning efficiency and improve radiotherapy plan quality. Current state-of-the-art techniques rely on convolutional neu...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Vasant Kearney, Jason W. Chan, Tianqi Wang, Alan Perry, Martina Descovich, Olivier Morin, Sue S. Yom, Timothy D. Solberg
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2020
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/1308bcc4ec654ecf9fb3ab5e9bc669a1
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:1308bcc4ec654ecf9fb3ab5e9bc669a1
record_format	dspace
spelling	oai:doaj.org-article:1308bcc4ec654ecf9fb3ab5e9bc669a12021-12-02T15:39:58ZDoseGAN: a generative adversarial network for synthetic dose prediction using attention-gated discrimination and generation10.1038/s41598-020-68062-72045-2322https://doaj.org/article/1308bcc4ec654ecf9fb3ab5e9bc669a12020-07-01T00:00:00Zhttps://doi.org/10.1038/s41598-020-68062-7https://doaj.org/toc/2045-2322Abstract Deep learning algorithms have recently been developed that utilize patient anatomy and raw imaging information to predict radiation dose, as a means to increase treatment planning efficiency and improve radiotherapy plan quality. Current state-of-the-art techniques rely on convolutional neural networks (CNNs) that use pixel-to-pixel loss to update network parameters. However, stereotactic body radiotherapy (SBRT) dose is often heterogeneous, making it difficult to model using pixel-level loss. Generative adversarial networks (GANs) utilize adversarial learning that incorporates image-level loss and is better suited to learn from heterogeneous labels. However, GANs are difficult to train and rely on compromised architectures to facilitate convergence. This study suggests an attention-gated generative adversarial network (DoseGAN) to improve learning, increase model complexity, and reduce network redundancy by focusing on relevant anatomy. DoseGAN was compared to alternative state-of-the-art dose prediction algorithms using heterogeneity index, conformity index, and various dosimetric parameters. All algorithms were trained, validated, and tested using 141 prostate SBRT patients. DoseGAN was able to predict more realistic volumetric dosimetry compared to all other algorithms and achieved statistically significant improvement compared to all alternative algorithms for the V100 and V120 of the PTV, V60 of the rectum, and heterogeneity index.Vasant KearneyJason W. ChanTianqi WangAlan PerryMartina DescovichOlivier MorinSue S. YomTimothy D. SolbergNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 10, Iss 1, Pp 1-8 (2020)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Vasant Kearney Jason W. Chan Tianqi Wang Alan Perry Martina Descovich Olivier Morin Sue S. Yom Timothy D. Solberg DoseGAN: a generative adversarial network for synthetic dose prediction using attention-gated discrimination and generation
description	Abstract Deep learning algorithms have recently been developed that utilize patient anatomy and raw imaging information to predict radiation dose, as a means to increase treatment planning efficiency and improve radiotherapy plan quality. Current state-of-the-art techniques rely on convolutional neural networks (CNNs) that use pixel-to-pixel loss to update network parameters. However, stereotactic body radiotherapy (SBRT) dose is often heterogeneous, making it difficult to model using pixel-level loss. Generative adversarial networks (GANs) utilize adversarial learning that incorporates image-level loss and is better suited to learn from heterogeneous labels. However, GANs are difficult to train and rely on compromised architectures to facilitate convergence. This study suggests an attention-gated generative adversarial network (DoseGAN) to improve learning, increase model complexity, and reduce network redundancy by focusing on relevant anatomy. DoseGAN was compared to alternative state-of-the-art dose prediction algorithms using heterogeneity index, conformity index, and various dosimetric parameters. All algorithms were trained, validated, and tested using 141 prostate SBRT patients. DoseGAN was able to predict more realistic volumetric dosimetry compared to all other algorithms and achieved statistically significant improvement compared to all alternative algorithms for the V100 and V120 of the PTV, V60 of the rectum, and heterogeneity index.
format	article
author	Vasant Kearney Jason W. Chan Tianqi Wang Alan Perry Martina Descovich Olivier Morin Sue S. Yom Timothy D. Solberg
author_facet	Vasant Kearney Jason W. Chan Tianqi Wang Alan Perry Martina Descovich Olivier Morin Sue S. Yom Timothy D. Solberg
author_sort	Vasant Kearney
title	DoseGAN: a generative adversarial network for synthetic dose prediction using attention-gated discrimination and generation
title_short	DoseGAN: a generative adversarial network for synthetic dose prediction using attention-gated discrimination and generation
title_full	DoseGAN: a generative adversarial network for synthetic dose prediction using attention-gated discrimination and generation
title_fullStr	DoseGAN: a generative adversarial network for synthetic dose prediction using attention-gated discrimination and generation
title_full_unstemmed	DoseGAN: a generative adversarial network for synthetic dose prediction using attention-gated discrimination and generation
title_sort	dosegan: a generative adversarial network for synthetic dose prediction using attention-gated discrimination and generation
publisher	Nature Portfolio
publishDate	2020
url	https://doaj.org/article/1308bcc4ec654ecf9fb3ab5e9bc669a1
work_keys_str_mv	AT vasantkearney doseganagenerativeadversarialnetworkforsyntheticdosepredictionusingattentiongateddiscriminationandgeneration AT jasonwchan doseganagenerativeadversarialnetworkforsyntheticdosepredictionusingattentiongateddiscriminationandgeneration AT tianqiwang doseganagenerativeadversarialnetworkforsyntheticdosepredictionusingattentiongateddiscriminationandgeneration AT alanperry doseganagenerativeadversarialnetworkforsyntheticdosepredictionusingattentiongateddiscriminationandgeneration AT martinadescovich doseganagenerativeadversarialnetworkforsyntheticdosepredictionusingattentiongateddiscriminationandgeneration AT oliviermorin doseganagenerativeadversarialnetworkforsyntheticdosepredictionusingattentiongateddiscriminationandgeneration AT suesyom doseganagenerativeadversarialnetworkforsyntheticdosepredictionusingattentiongateddiscriminationandgeneration AT timothydsolberg doseganagenerativeadversarialnetworkforsyntheticdosepredictionusingattentiongateddiscriminationandgeneration
_version_	1718385855407587328

DoseGAN: a generative adversarial network for synthetic dose prediction using attention-gated discrimination and generation

Ejemplares similares