Novel radiotherapy target definition using AI-driven predictions of glioblastoma recurrence from metabolic and diffusion MRI

Abstract The current standard-of-care (SOC) practice for defining the clinical target volume (CTV) for radiation therapy (RT) in patients with glioblastoma still employs an isotropic 1–2 cm expansion of the T2-hyperintensity lesion, without considering the heterogeneous infiltrative nature of these...

Full description

Saved in:
Bibliographic Details
Main Authors: Nate Tran, Tracy L. Luks, Yan Li, Angela Jakary, Jacob Ellison, Bo Liu, Oluwaseun Adegbite, Devika Nair, Pranav Kakhandiki, Annette M. Molinaro, Javier E. Villanueva-Meyer, Nicholas Butowski, Jennifer L. Clarke, Susan M. Chang, Steve E. Braunstein, Olivier Morin, Hui Lin, Janine M. Lupo
Format: Article
Language:English
Published: Nature Portfolio 2025-08-01
Series:npj Digital Medicine
Online Access:https://doi.org/10.1038/s41746-025-01861-2
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849341683613827072
author Nate Tran
Tracy L. Luks
Yan Li
Angela Jakary
Jacob Ellison
Bo Liu
Oluwaseun Adegbite
Devika Nair
Pranav Kakhandiki
Annette M. Molinaro
Javier E. Villanueva-Meyer
Nicholas Butowski
Jennifer L. Clarke
Susan M. Chang
Steve E. Braunstein
Olivier Morin
Hui Lin
Janine M. Lupo
author_facet Nate Tran
Tracy L. Luks
Yan Li
Angela Jakary
Jacob Ellison
Bo Liu
Oluwaseun Adegbite
Devika Nair
Pranav Kakhandiki
Annette M. Molinaro
Javier E. Villanueva-Meyer
Nicholas Butowski
Jennifer L. Clarke
Susan M. Chang
Steve E. Braunstein
Olivier Morin
Hui Lin
Janine M. Lupo
author_sort Nate Tran
collection DOAJ
description Abstract The current standard-of-care (SOC) practice for defining the clinical target volume (CTV) for radiation therapy (RT) in patients with glioblastoma still employs an isotropic 1–2 cm expansion of the T2-hyperintensity lesion, without considering the heterogeneous infiltrative nature of these tumors. This study aims to improve RT CTV definition in patients with glioblastoma by incorporating biologically relevant metabolic and physiologic imaging acquired before RT along with a deep learning model that can predict regions of subsequent tumor progression by either the presence of contrast-enhancement or T2-hyperintensity. The results were compared against two standard CTV definitions. Our multi-parametric deep learning model significantly outperformed the uniform 2 cm expansion of the T2-lesion CTV in terms of specificity (0.89 ± 0.05 vs 0.79 ± 0.11; p = 0.004), while also achieving comparable sensitivity (0.92 ± 0.11 vs 0.95 ± 0.08; p = 0.10), sparing more normal brain. Model performance was significantly enhanced by incorporating lesion size-weighted loss functions during training and including metabolic images as inputs.
format Article
id doaj-art-b5abaec0a4e24f35912a3e464393975d
institution Kabale University
issn 2398-6352
language English
publishDate 2025-08-01
publisher Nature Portfolio
record_format Article
series npj Digital Medicine
spelling doaj-art-b5abaec0a4e24f35912a3e464393975d2025-08-20T03:43:34ZengNature Portfolionpj Digital Medicine2398-63522025-08-018111210.1038/s41746-025-01861-2Novel radiotherapy target definition using AI-driven predictions of glioblastoma recurrence from metabolic and diffusion MRINate Tran0Tracy L. Luks1Yan Li2Angela Jakary3Jacob Ellison4Bo Liu5Oluwaseun Adegbite6Devika Nair7Pranav Kakhandiki8Annette M. Molinaro9Javier E. Villanueva-Meyer10Nicholas Butowski11Jennifer L. Clarke12Susan M. Chang13Steve E. Braunstein14Olivier Morin15Hui Lin16Janine M. Lupo17Department of Radiology & Biomedical Imaging, University of CaliforniaDepartment of Radiology & Biomedical Imaging, University of CaliforniaDepartment of Radiology & Biomedical Imaging, University of CaliforniaDepartment of Radiology & Biomedical Imaging, University of CaliforniaDepartment of Radiology & Biomedical Imaging, University of CaliforniaDepartment of Radiology & Biomedical Imaging, University of CaliforniaDepartment of Radiology & Biomedical Imaging, University of CaliforniaDepartment of Radiology & Biomedical Imaging, University of CaliforniaDepartment of Radiology & Biomedical Imaging, University of CaliforniaDepartment of Neurological Surgery, University of CaliforniaDepartment of Radiology & Biomedical Imaging, University of CaliforniaDepartment of Neurological Surgery, University of CaliforniaDepartment of Neurological Surgery, University of CaliforniaDepartment of Neurological Surgery, University of CaliforniaDepartment of Radiation Oncology, University of CaliforniaDepartment of Radiation Oncology, University of CaliforniaUCSF/UC Berkeley Graduate Program in Bioengineering, University of CaliforniaDepartment of Radiology & Biomedical Imaging, University of CaliforniaAbstract The current standard-of-care (SOC) practice for defining the clinical target volume (CTV) for radiation therapy (RT) in patients with glioblastoma still employs an isotropic 1–2 cm expansion of the T2-hyperintensity lesion, without considering the heterogeneous infiltrative nature of these tumors. This study aims to improve RT CTV definition in patients with glioblastoma by incorporating biologically relevant metabolic and physiologic imaging acquired before RT along with a deep learning model that can predict regions of subsequent tumor progression by either the presence of contrast-enhancement or T2-hyperintensity. The results were compared against two standard CTV definitions. Our multi-parametric deep learning model significantly outperformed the uniform 2 cm expansion of the T2-lesion CTV in terms of specificity (0.89 ± 0.05 vs 0.79 ± 0.11; p = 0.004), while also achieving comparable sensitivity (0.92 ± 0.11 vs 0.95 ± 0.08; p = 0.10), sparing more normal brain. Model performance was significantly enhanced by incorporating lesion size-weighted loss functions during training and including metabolic images as inputs.https://doi.org/10.1038/s41746-025-01861-2
spellingShingle Nate Tran
Tracy L. Luks
Yan Li
Angela Jakary
Jacob Ellison
Bo Liu
Oluwaseun Adegbite
Devika Nair
Pranav Kakhandiki
Annette M. Molinaro
Javier E. Villanueva-Meyer
Nicholas Butowski
Jennifer L. Clarke
Susan M. Chang
Steve E. Braunstein
Olivier Morin
Hui Lin
Janine M. Lupo
Novel radiotherapy target definition using AI-driven predictions of glioblastoma recurrence from metabolic and diffusion MRI
npj Digital Medicine
title Novel radiotherapy target definition using AI-driven predictions of glioblastoma recurrence from metabolic and diffusion MRI
title_full Novel radiotherapy target definition using AI-driven predictions of glioblastoma recurrence from metabolic and diffusion MRI
title_fullStr Novel radiotherapy target definition using AI-driven predictions of glioblastoma recurrence from metabolic and diffusion MRI
title_full_unstemmed Novel radiotherapy target definition using AI-driven predictions of glioblastoma recurrence from metabolic and diffusion MRI
title_short Novel radiotherapy target definition using AI-driven predictions of glioblastoma recurrence from metabolic and diffusion MRI
title_sort novel radiotherapy target definition using ai driven predictions of glioblastoma recurrence from metabolic and diffusion mri
url https://doi.org/10.1038/s41746-025-01861-2
work_keys_str_mv AT natetran novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT tracylluks novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT yanli novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT angelajakary novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT jacobellison novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT boliu novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT oluwaseunadegbite novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT devikanair novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT pranavkakhandiki novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT annettemmolinaro novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT javierevillanuevameyer novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT nicholasbutowski novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT jenniferlclarke novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT susanmchang novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT steveebraunstein novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT oliviermorin novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT huilin novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri
AT janinemlupo novelradiotherapytargetdefinitionusingaidrivenpredictionsofglioblastomarecurrencefrommetabolicanddiffusionmri