Glioblastoma Surgery Imaging–Reporting and Data System: Validation and Performance of the Automated Segmentation Task
For patients with presumed glioblastoma, essential tumor characteristics are determined from preoperative MR images to optimize the treatment strategy. This procedure is time-consuming and subjective, if performed by crude eyeballing or manually. The standardized GSI-RADS aims to provide neurosurgeo...
Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-09-01
|
Series: | Cancers |
Subjects: | |
Online Access: | https://www.mdpi.com/2072-6694/13/18/4674 |
_version_ | 1797519980362727424 |
---|---|
author | David Bouget Roelant S. Eijgelaar André Pedersen Ivar Kommers Hilko Ardon Frederik Barkhof Lorenzo Bello Mitchel S. Berger Marco Conti Nibali Julia Furtner Even Hovig Fyllingen Shawn Hervey-Jumper Albert J. S. Idema Barbara Kiesel Alfred Kloet Emmanuel Mandonnet Domenique M. J. Müller Pierre A. Robe Marco Rossi Lisa M. Sagberg Tommaso Sciortino Wimar A. Van den Brink Michiel Wagemakers Georg Widhalm Marnix G. Witte Aeilko H. Zwinderman Ingerid Reinertsen Philip C. De Witt Hamer Ole Solheim |
author_facet | David Bouget Roelant S. Eijgelaar André Pedersen Ivar Kommers Hilko Ardon Frederik Barkhof Lorenzo Bello Mitchel S. Berger Marco Conti Nibali Julia Furtner Even Hovig Fyllingen Shawn Hervey-Jumper Albert J. S. Idema Barbara Kiesel Alfred Kloet Emmanuel Mandonnet Domenique M. J. Müller Pierre A. Robe Marco Rossi Lisa M. Sagberg Tommaso Sciortino Wimar A. Van den Brink Michiel Wagemakers Georg Widhalm Marnix G. Witte Aeilko H. Zwinderman Ingerid Reinertsen Philip C. De Witt Hamer Ole Solheim |
author_sort | David Bouget |
collection | DOAJ |
description | For patients with presumed glioblastoma, essential tumor characteristics are determined from preoperative MR images to optimize the treatment strategy. This procedure is time-consuming and subjective, if performed by crude eyeballing or manually. The standardized GSI-RADS aims to provide neurosurgeons with automatic tumor segmentations to extract tumor features rapidly and objectively. In this study, we improved automatic tumor segmentation and compared the agreement with manual raters, describe the technical details of the different components of GSI-RADS, and determined their speed. Two recent neural network architectures were considered for the segmentation task: nnU-Net and AGU-Net. Two preprocessing schemes were introduced to investigate the tradeoff between performance and processing speed. A summarized description of the tumor feature extraction and standardized reporting process is included. The trained architectures for automatic segmentation and the code for computing the standardized report are distributed as open-source and as open-access software. Validation studies were performed on a dataset of 1594 gadolinium-enhanced T1-weighted MRI volumes from 13 hospitals and 293 T1-weighted MRI volumes from the BraTS challenge. The glioblastoma tumor core segmentation reached a Dice score slightly below 90%, a patientwise F1-score close to 99%, and a 95th percentile Hausdorff distance slightly below <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>4.0</mn></mrow></semantics></math></inline-formula> mm on average with either architecture and the heavy preprocessing scheme. A patient MRI volume can be segmented in less than one minute, and a standardized report can be generated in up to five minutes. The proposed GSI-RADS software showed robust performance on a large collection of MRI volumes from various hospitals and generated results within a reasonable runtime. |
first_indexed | 2024-03-10T07:50:23Z |
format | Article |
id | doaj.art-7113f57916744c6fab4844515bad6eab |
institution | Directory Open Access Journal |
issn | 2072-6694 |
language | English |
last_indexed | 2024-03-10T07:50:23Z |
publishDate | 2021-09-01 |
publisher | MDPI AG |
record_format | Article |
series | Cancers |
spelling | doaj.art-7113f57916744c6fab4844515bad6eab2023-11-22T12:18:50ZengMDPI AGCancers2072-66942021-09-011318467410.3390/cancers13184674Glioblastoma Surgery Imaging–Reporting and Data System: Validation and Performance of the Automated Segmentation TaskDavid Bouget0Roelant S. Eijgelaar1André Pedersen2Ivar Kommers3Hilko Ardon4Frederik Barkhof5Lorenzo Bello6Mitchel S. Berger7Marco Conti Nibali8Julia Furtner9Even Hovig Fyllingen10Shawn Hervey-Jumper11Albert J. S. Idema12Barbara Kiesel13Alfred Kloet14Emmanuel Mandonnet15Domenique M. J. Müller16Pierre A. Robe17Marco Rossi18Lisa M. Sagberg19Tommaso Sciortino20Wimar A. Van den Brink21Michiel Wagemakers22Georg Widhalm23Marnix G. Witte24Aeilko H. Zwinderman25Ingerid Reinertsen26Philip C. De Witt Hamer27Ole Solheim28Department of Health Research, SINTEF Digital, NO-7465 Trondheim, NorwayDepartment of Neurosurgery, Amsterdam University Medical Centers, Vrije Universiteit, 1081 HV Amsterdam, The NetherlandsDepartment of Health Research, SINTEF Digital, NO-7465 Trondheim, NorwayDepartment of Neurosurgery, Amsterdam University Medical Centers, Vrije Universiteit, 1081 HV Amsterdam, The NetherlandsDepartment of Neurosurgery, Twee Steden Hospital, 5042 AD Tilburg, The NetherlandsDepartment of Radiology and Nuclear Medicine, Amsterdam University Medical Centers, Vrije Universiteit, 1081 HV Amsterdam, The NetherlandsNeurosurgical Oncology Unit, Department of Oncology and Hemato-Oncology, Humanitas Research Hospital, Università Degli Studi di Milano, 20122 Milano, ItalyDepartment of Neurological Surgery, University of California, San Francisco, CA 94143, USANeurosurgical Oncology Unit, Department of Oncology and Hemato-Oncology, Humanitas Research Hospital, Università Degli Studi di Milano, 20122 Milano, ItalyDepartment of Biomedical Imaging and Image-Guided Therapy, Medical University Vienna, 1090 Wien, AustriaDepartment of Circulation and Medical Imaging, Norwegian University of Science and Technology, NO-7491 Trondheim, NorwayDepartment of Neurological Surgery, University of California, San Francisco, CA 94143, USADepartment of Neurosurgery, Northwest Clinics, 1815 JD Alkmaar, The NetherlandsDepartment of Neurosurgery, Medical University Vienna, 1090 Wien, AustriaDepartment of Neurosurgery, Haaglanden Medical Center, 2512 VA The Hague, The NetherlandsDepartment of Neurological Surgery, Hôpital Lariboisière, 75010 Paris, FranceDepartment of Neurosurgery, Amsterdam University Medical Centers, Vrije Universiteit, 1081 HV Amsterdam, The NetherlandsDepartment of Neurology and Neurosurgery, University Medical Center Utrecht, 3584 CX Utrecht, The NetherlandsNeurosurgical Oncology Unit, Department of Oncology and Hemato-Oncology, Humanitas Research Hospital, Università Degli Studi di Milano, 20122 Milano, ItalyDepartment of Neurosurgery, St. Olavs Hospital, Trondheim University Hospital, NO-7030 Trondheim, NorwayNeurosurgical Oncology Unit, Department of Oncology and Hemato-Oncology, Humanitas Research Hospital, Università Degli Studi di Milano, 20122 Milano, ItalyDepartment of Neurosurgery, Isala Hospital Zwolle, 8025 AB Zwolle, The NetherlandsDepartment of Neurosurgery, University Medical Center Groningen, University of Groningen, 9713 GZ Groningen, The NetherlandsDepartment of Neurosurgery, Medical University Vienna, 1090 Wien, AustriaDepartment of Radiation Oncology, The Netherlands Cancer Institute, 1066 CX Amsterdam, The NetherlandsDepartment of Clinical Epidemiology and Biostatistics, Amsterdam University Medical Centers, 1105 AZ Amsterdam, The NetherlandsDepartment of Health Research, SINTEF Digital, NO-7465 Trondheim, NorwayDepartment of Neurosurgery, Amsterdam University Medical Centers, Vrije Universiteit, 1081 HV Amsterdam, The NetherlandsDepartment of Clinical Epidemiology and Biostatistics, Amsterdam University Medical Centers, 1105 AZ Amsterdam, The NetherlandsFor patients with presumed glioblastoma, essential tumor characteristics are determined from preoperative MR images to optimize the treatment strategy. This procedure is time-consuming and subjective, if performed by crude eyeballing or manually. The standardized GSI-RADS aims to provide neurosurgeons with automatic tumor segmentations to extract tumor features rapidly and objectively. In this study, we improved automatic tumor segmentation and compared the agreement with manual raters, describe the technical details of the different components of GSI-RADS, and determined their speed. Two recent neural network architectures were considered for the segmentation task: nnU-Net and AGU-Net. Two preprocessing schemes were introduced to investigate the tradeoff between performance and processing speed. A summarized description of the tumor feature extraction and standardized reporting process is included. The trained architectures for automatic segmentation and the code for computing the standardized report are distributed as open-source and as open-access software. Validation studies were performed on a dataset of 1594 gadolinium-enhanced T1-weighted MRI volumes from 13 hospitals and 293 T1-weighted MRI volumes from the BraTS challenge. The glioblastoma tumor core segmentation reached a Dice score slightly below 90%, a patientwise F1-score close to 99%, and a 95th percentile Hausdorff distance slightly below <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><mn>4.0</mn></mrow></semantics></math></inline-formula> mm on average with either architecture and the heavy preprocessing scheme. A patient MRI volume can be segmented in less than one minute, and a standardized report can be generated in up to five minutes. The proposed GSI-RADS software showed robust performance on a large collection of MRI volumes from various hospitals and generated results within a reasonable runtime.https://www.mdpi.com/2072-6694/13/18/4674glioblastomadeep learning3D segmentationcomputer-assisted image processingmagnetic resonance imagingneuroimaging |
spellingShingle | David Bouget Roelant S. Eijgelaar André Pedersen Ivar Kommers Hilko Ardon Frederik Barkhof Lorenzo Bello Mitchel S. Berger Marco Conti Nibali Julia Furtner Even Hovig Fyllingen Shawn Hervey-Jumper Albert J. S. Idema Barbara Kiesel Alfred Kloet Emmanuel Mandonnet Domenique M. J. Müller Pierre A. Robe Marco Rossi Lisa M. Sagberg Tommaso Sciortino Wimar A. Van den Brink Michiel Wagemakers Georg Widhalm Marnix G. Witte Aeilko H. Zwinderman Ingerid Reinertsen Philip C. De Witt Hamer Ole Solheim Glioblastoma Surgery Imaging–Reporting and Data System: Validation and Performance of the Automated Segmentation Task Cancers glioblastoma deep learning 3D segmentation computer-assisted image processing magnetic resonance imaging neuroimaging |
title | Glioblastoma Surgery Imaging–Reporting and Data System: Validation and Performance of the Automated Segmentation Task |
title_full | Glioblastoma Surgery Imaging–Reporting and Data System: Validation and Performance of the Automated Segmentation Task |
title_fullStr | Glioblastoma Surgery Imaging–Reporting and Data System: Validation and Performance of the Automated Segmentation Task |
title_full_unstemmed | Glioblastoma Surgery Imaging–Reporting and Data System: Validation and Performance of the Automated Segmentation Task |
title_short | Glioblastoma Surgery Imaging–Reporting and Data System: Validation and Performance of the Automated Segmentation Task |
title_sort | glioblastoma surgery imaging reporting and data system validation and performance of the automated segmentation task |
topic | glioblastoma deep learning 3D segmentation computer-assisted image processing magnetic resonance imaging neuroimaging |
url | https://www.mdpi.com/2072-6694/13/18/4674 |
work_keys_str_mv | AT davidbouget glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT roelantseijgelaar glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT andrepedersen glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT ivarkommers glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT hilkoardon glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT frederikbarkhof glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT lorenzobello glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT mitchelsberger glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT marcocontinibali glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT juliafurtner glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT evenhovigfyllingen glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT shawnherveyjumper glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT albertjsidema glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT barbarakiesel glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT alfredkloet glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT emmanuelmandonnet glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT domeniquemjmuller glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT pierrearobe glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT marcorossi glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT lisamsagberg glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT tommasosciortino glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT wimaravandenbrink glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT michielwagemakers glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT georgwidhalm glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT marnixgwitte glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT aeilkohzwinderman glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT ingeridreinertsen glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT philipcdewitthamer glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask AT olesolheim glioblastomasurgeryimagingreportinganddatasystemvalidationandperformanceoftheautomatedsegmentationtask |