Validation of SynthSeg segmentation performance on CT using paired MRI from radiotherapy patients

Introduction:: Manual segmentation of medical images is labor intensive and especially challenging for images with poor contrast or resolution. The presence of disease exacerbates this further, increasing the need for an automated solution. To this extent, SynthSeg is a robust deep learning model de...

Full description

Saved in:
Bibliographic Details
Main Authors: Selena Huisman, Matteo Maspero, Marielle Philippens, Joost Verhoeff, Szabolcs David
Format: Article
Language:English
Published: Elsevier 2024-12-01
Series:NeuroImage
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S1053811924004191
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846150189715292160
author Selena Huisman
Matteo Maspero
Marielle Philippens
Joost Verhoeff
Szabolcs David
author_facet Selena Huisman
Matteo Maspero
Marielle Philippens
Joost Verhoeff
Szabolcs David
author_sort Selena Huisman
collection DOAJ
description Introduction:: Manual segmentation of medical images is labor intensive and especially challenging for images with poor contrast or resolution. The presence of disease exacerbates this further, increasing the need for an automated solution. To this extent, SynthSeg is a robust deep learning model designed for automatic brain segmentation across various contrasts and resolutions. This study validates the SynthSeg robust brain segmentation model on computed tomography (CT), using a multi-center dataset. Methods:: An open access dataset of 260 paired CT and magnetic resonance imaging (MRI) from radiotherapy patients treated in 5 centers was collected. Brain segmentations from CT and MRI were obtained with SynthSeg model, a component of the Freesurfer imaging suite. These segmentations were compared and evaluated using Dice scores and Hausdorff 95 distance (HD95), treating MRI-based segmentations as the ground truth. Brain regions that failed to meet performance criteria were excluded based on automated quality control (QC) scores. Results:: Dice scores indicate a median overlap of 0.76 (IQR: 0.65-0.83). The mean volume difference is 7.79% (CI: 6.41%–9.18%), with CT segmentations typically smaller than MRI-based. The median HD95 is 2.95 mm (IQR: 1.73-5.39). QC score based thresholding improves median dice by 0.1 and median HD95 by 0.05 mm. Morphological differences related to sex and age, as detected by MRI, were also replicated with CT, with an approximate 17% difference between the CT and MRI results for sex and 10% difference between the results for age. Conclusion:: SynthSeg can be utilized for CT-based automatic brain segmentation, but only in applications where precision is not essential. CT performance is lower than MRI based on the integrated QC scores, but low-quality segmentations can be excluded with QC-based thresholding. Additionally, performing CT-based neuroanatomical studies is encouraged, as the results show correlations in sex- and age-based analyses similar to those found with MRI.
format Article
id doaj-art-a8bd83ce9a434a7e98d6163b4117d65b
institution Kabale University
issn 1095-9572
language English
publishDate 2024-12-01
publisher Elsevier
record_format Article
series NeuroImage
spelling doaj-art-a8bd83ce9a434a7e98d6163b4117d65b2024-11-29T06:22:59ZengElsevierNeuroImage1095-95722024-12-01303120922Validation of SynthSeg segmentation performance on CT using paired MRI from radiotherapy patientsSelena Huisman0Matteo Maspero1Marielle Philippens2Joost Verhoeff3Szabolcs David4Department of Radiation Oncology, Amsterdam UMC, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands; Department of Radiation Oncology, UMC Utrecht, Heidelberglaan 100, 3508 GA Utrecht, The Netherlands; Corresponding author at: Department of Radiation Oncology, Amsterdam UMC, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands.Department of Radiation Oncology, UMC Utrecht, Heidelberglaan 100, 3508 GA Utrecht, The Netherlands; Computational Imaging Group for MR Diagnostics & Therapy, UMC Utrecht, Heidelberglaan 100, 3508 GA Utrecht, The NetherlandsDepartment of Radiation Oncology, UMC Utrecht, Heidelberglaan 100, 3508 GA Utrecht, The NetherlandsDepartment of Radiation Oncology, Amsterdam UMC, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands; Department of Radiation Oncology, UMC Utrecht, Heidelberglaan 100, 3508 GA Utrecht, The NetherlandsDepartment of Radiation Oncology, Amsterdam UMC, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands; Department of Radiation Oncology, UMC Utrecht, Heidelberglaan 100, 3508 GA Utrecht, The NetherlandsIntroduction:: Manual segmentation of medical images is labor intensive and especially challenging for images with poor contrast or resolution. The presence of disease exacerbates this further, increasing the need for an automated solution. To this extent, SynthSeg is a robust deep learning model designed for automatic brain segmentation across various contrasts and resolutions. This study validates the SynthSeg robust brain segmentation model on computed tomography (CT), using a multi-center dataset. Methods:: An open access dataset of 260 paired CT and magnetic resonance imaging (MRI) from radiotherapy patients treated in 5 centers was collected. Brain segmentations from CT and MRI were obtained with SynthSeg model, a component of the Freesurfer imaging suite. These segmentations were compared and evaluated using Dice scores and Hausdorff 95 distance (HD95), treating MRI-based segmentations as the ground truth. Brain regions that failed to meet performance criteria were excluded based on automated quality control (QC) scores. Results:: Dice scores indicate a median overlap of 0.76 (IQR: 0.65-0.83). The mean volume difference is 7.79% (CI: 6.41%–9.18%), with CT segmentations typically smaller than MRI-based. The median HD95 is 2.95 mm (IQR: 1.73-5.39). QC score based thresholding improves median dice by 0.1 and median HD95 by 0.05 mm. Morphological differences related to sex and age, as detected by MRI, were also replicated with CT, with an approximate 17% difference between the CT and MRI results for sex and 10% difference between the results for age. Conclusion:: SynthSeg can be utilized for CT-based automatic brain segmentation, but only in applications where precision is not essential. CT performance is lower than MRI based on the integrated QC scores, but low-quality segmentations can be excluded with QC-based thresholding. Additionally, performing CT-based neuroanatomical studies is encouraged, as the results show correlations in sex- and age-based analyses similar to those found with MRI.http://www.sciencedirect.com/science/article/pii/S1053811924004191Deep learningValidationClinical brain MRIClinical brain CT
spellingShingle Selena Huisman
Matteo Maspero
Marielle Philippens
Joost Verhoeff
Szabolcs David
Validation of SynthSeg segmentation performance on CT using paired MRI from radiotherapy patients
NeuroImage
Deep learning
Validation
Clinical brain MRI
Clinical brain CT
title Validation of SynthSeg segmentation performance on CT using paired MRI from radiotherapy patients
title_full Validation of SynthSeg segmentation performance on CT using paired MRI from radiotherapy patients
title_fullStr Validation of SynthSeg segmentation performance on CT using paired MRI from radiotherapy patients
title_full_unstemmed Validation of SynthSeg segmentation performance on CT using paired MRI from radiotherapy patients
title_short Validation of SynthSeg segmentation performance on CT using paired MRI from radiotherapy patients
title_sort validation of synthseg segmentation performance on ct using paired mri from radiotherapy patients
topic Deep learning
Validation
Clinical brain MRI
Clinical brain CT
url http://www.sciencedirect.com/science/article/pii/S1053811924004191
work_keys_str_mv AT selenahuisman validationofsynthsegsegmentationperformanceonctusingpairedmrifromradiotherapypatients
AT matteomaspero validationofsynthsegsegmentationperformanceonctusingpairedmrifromradiotherapypatients
AT mariellephilippens validationofsynthsegsegmentationperformanceonctusingpairedmrifromradiotherapypatients
AT joostverhoeff validationofsynthsegsegmentationperformanceonctusingpairedmrifromradiotherapypatients
AT szabolcsdavid validationofsynthsegsegmentationperformanceonctusingpairedmrifromradiotherapypatients