Fully automated deep learning based auto-contouring of liver segments and spleen on contrast-enhanced CT images
Abstract Manual delineation of liver segments on computed tomography (CT) images for primary/secondary liver cancer (LC) patients is time-intensive and prone to inter/intra-observer variability. Therefore, we developed a deep-learning-based model to auto-contour liver segments and spleen on contrast...
Main Authors: | , , , , , , , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Nature Portfolio
2024-02-01
|
Series: | Scientific Reports |
Online Access: | https://doi.org/10.1038/s41598-024-53997-y |
_version_ | 1797275011879272448 |
---|---|
author | Aashish C. Gupta Guillaume Cazoulat Mais Al Taie Sireesha Yedururi Bastien Rigaud Austin Castelo John Wood Cenji Yu Caleb O’Connor Usama Salem Jessica Albuquerque Marques Silva Aaron Kyle Jones Molly McCulloch Bruno C. Odisio Eugene J. Koay Kristy K. Brock |
author_facet | Aashish C. Gupta Guillaume Cazoulat Mais Al Taie Sireesha Yedururi Bastien Rigaud Austin Castelo John Wood Cenji Yu Caleb O’Connor Usama Salem Jessica Albuquerque Marques Silva Aaron Kyle Jones Molly McCulloch Bruno C. Odisio Eugene J. Koay Kristy K. Brock |
author_sort | Aashish C. Gupta |
collection | DOAJ |
description | Abstract Manual delineation of liver segments on computed tomography (CT) images for primary/secondary liver cancer (LC) patients is time-intensive and prone to inter/intra-observer variability. Therefore, we developed a deep-learning-based model to auto-contour liver segments and spleen on contrast-enhanced CT (CECT) images. We trained two models using 3d patch-based attention U-Net ( $${{\text{M}}}_{{\text{paU}}-{\text{Net}}})$$ M paU - Net ) and 3d full resolution of nnU-Net ( $${{\text{M}}}_{{\text{nnU}}-{\text{Net}}})$$ M nnU - Net ) to determine the best architecture ( $${\text{BA}})$$ BA ) . BA was used with vessels ( $${{\text{M}}}_{{\text{Vess}}})$$ M Vess ) and spleen ( $${{\text{M}}}_{{\text{seg}}+{\text{spleen}}})$$ M seg + spleen ) to assess the impact on segment contouring. Models were trained, validated, and tested on 160 ( $${{\text{C}}}_{{\text{RTTrain}}}$$ C RTTrain ), 40 ( $${{\text{C}}}_{{\text{RTVal}}}$$ C RTVal ), 33 ( $${{\text{C}}}_{{\text{LS}}}$$ C LS ), 25 (CCH) and 20 (CPVE) CECT of LC patients. $${{\text{M}}}_{{\text{nnU}}-{\text{Net}}}$$ M nnU - Net outperformed $${{\text{M}}}_{{\text{paU}}-{\text{Net}}}$$ M paU - Net across all segments with median differences in Dice similarity coefficients (DSC) ranging 0.03–0.05 (p < 0.05). $${{\text{M}}}_{{\text{seg}}+{\text{spleen}}}$$ M seg + spleen , and $${{\text{M}}}_{{\text{nnU}}-{\text{Net}}}$$ M nnU - Net were not statistically different (p > 0.05), however, both were slightly better than $${{\text{M}}}_{{\text{Vess}}}$$ M Vess by DSC up to 0.02. The final model, $${{\text{M}}}_{{\text{seg}}+{\text{spleen}}}$$ M seg + spleen , showed a mean DSC of 0.89, 0.82, 0.88, 0.87, 0.96, and 0.95 for segments 1, 2, 3, 4, 5–8, and spleen, respectively on entire test sets. Qualitatively, more than 85% of cases showed a Likert score $$\ge$$ ≥ 3 on test sets. Our final model provides clinically acceptable contours of liver segments and spleen which are usable in treatment planning. |
first_indexed | 2024-03-07T15:08:14Z |
format | Article |
id | doaj.art-6f8b71221bc04ad193b13173d4a5714c |
institution | Directory Open Access Journal |
issn | 2045-2322 |
language | English |
last_indexed | 2024-03-07T15:08:14Z |
publishDate | 2024-02-01 |
publisher | Nature Portfolio |
record_format | Article |
series | Scientific Reports |
spelling | doaj.art-6f8b71221bc04ad193b13173d4a5714c2024-03-05T18:48:42ZengNature PortfolioScientific Reports2045-23222024-02-0114111910.1038/s41598-024-53997-yFully automated deep learning based auto-contouring of liver segments and spleen on contrast-enhanced CT imagesAashish C. Gupta0Guillaume Cazoulat1Mais Al Taie2Sireesha Yedururi3Bastien Rigaud4Austin Castelo5John Wood6Cenji Yu7Caleb O’Connor8Usama Salem9Jessica Albuquerque Marques Silva10Aaron Kyle Jones11Molly McCulloch12Bruno C. Odisio13Eugene J. Koay14Kristy K. Brock15Department of Imaging Physics, The University of Texas MD Anderson Cancer CenterDepartment of Imaging Physics, The University of Texas MD Anderson Cancer CenterDepartment of Imaging Physics, The University of Texas MD Anderson Cancer CenterAbdominal Imaging Department, The University of Texas MD Anderson Cancer CenterDepartment of Imaging Physics, The University of Texas MD Anderson Cancer CenterDepartment of Imaging Physics, The University of Texas MD Anderson Cancer CenterDepartment of Imaging Physics, The University of Texas MD Anderson Cancer CenterThe University of Texas MD Anderson Cancer Center UTHealth Graduate School of Biomedical SciencesDepartment of Imaging Physics, The University of Texas MD Anderson Cancer CenterAbdominal Imaging Department, The University of Texas MD Anderson Cancer CenterDepartment of Interventional Radiology, The University of Texas MD Anderson Cancer CenterDepartment of Imaging Physics, The University of Texas MD Anderson Cancer CenterDepartment of Imaging Physics, The University of Texas MD Anderson Cancer CenterDepartment of Interventional Radiology, The University of Texas MD Anderson Cancer CenterThe University of Texas MD Anderson Cancer Center UTHealth Graduate School of Biomedical SciencesDepartment of Imaging Physics, The University of Texas MD Anderson Cancer CenterAbstract Manual delineation of liver segments on computed tomography (CT) images for primary/secondary liver cancer (LC) patients is time-intensive and prone to inter/intra-observer variability. Therefore, we developed a deep-learning-based model to auto-contour liver segments and spleen on contrast-enhanced CT (CECT) images. We trained two models using 3d patch-based attention U-Net ( $${{\text{M}}}_{{\text{paU}}-{\text{Net}}})$$ M paU - Net ) and 3d full resolution of nnU-Net ( $${{\text{M}}}_{{\text{nnU}}-{\text{Net}}})$$ M nnU - Net ) to determine the best architecture ( $${\text{BA}})$$ BA ) . BA was used with vessels ( $${{\text{M}}}_{{\text{Vess}}})$$ M Vess ) and spleen ( $${{\text{M}}}_{{\text{seg}}+{\text{spleen}}})$$ M seg + spleen ) to assess the impact on segment contouring. Models were trained, validated, and tested on 160 ( $${{\text{C}}}_{{\text{RTTrain}}}$$ C RTTrain ), 40 ( $${{\text{C}}}_{{\text{RTVal}}}$$ C RTVal ), 33 ( $${{\text{C}}}_{{\text{LS}}}$$ C LS ), 25 (CCH) and 20 (CPVE) CECT of LC patients. $${{\text{M}}}_{{\text{nnU}}-{\text{Net}}}$$ M nnU - Net outperformed $${{\text{M}}}_{{\text{paU}}-{\text{Net}}}$$ M paU - Net across all segments with median differences in Dice similarity coefficients (DSC) ranging 0.03–0.05 (p < 0.05). $${{\text{M}}}_{{\text{seg}}+{\text{spleen}}}$$ M seg + spleen , and $${{\text{M}}}_{{\text{nnU}}-{\text{Net}}}$$ M nnU - Net were not statistically different (p > 0.05), however, both were slightly better than $${{\text{M}}}_{{\text{Vess}}}$$ M Vess by DSC up to 0.02. The final model, $${{\text{M}}}_{{\text{seg}}+{\text{spleen}}}$$ M seg + spleen , showed a mean DSC of 0.89, 0.82, 0.88, 0.87, 0.96, and 0.95 for segments 1, 2, 3, 4, 5–8, and spleen, respectively on entire test sets. Qualitatively, more than 85% of cases showed a Likert score $$\ge$$ ≥ 3 on test sets. Our final model provides clinically acceptable contours of liver segments and spleen which are usable in treatment planning.https://doi.org/10.1038/s41598-024-53997-y |
spellingShingle | Aashish C. Gupta Guillaume Cazoulat Mais Al Taie Sireesha Yedururi Bastien Rigaud Austin Castelo John Wood Cenji Yu Caleb O’Connor Usama Salem Jessica Albuquerque Marques Silva Aaron Kyle Jones Molly McCulloch Bruno C. Odisio Eugene J. Koay Kristy K. Brock Fully automated deep learning based auto-contouring of liver segments and spleen on contrast-enhanced CT images Scientific Reports |
title | Fully automated deep learning based auto-contouring of liver segments and spleen on contrast-enhanced CT images |
title_full | Fully automated deep learning based auto-contouring of liver segments and spleen on contrast-enhanced CT images |
title_fullStr | Fully automated deep learning based auto-contouring of liver segments and spleen on contrast-enhanced CT images |
title_full_unstemmed | Fully automated deep learning based auto-contouring of liver segments and spleen on contrast-enhanced CT images |
title_short | Fully automated deep learning based auto-contouring of liver segments and spleen on contrast-enhanced CT images |
title_sort | fully automated deep learning based auto contouring of liver segments and spleen on contrast enhanced ct images |
url | https://doi.org/10.1038/s41598-024-53997-y |
work_keys_str_mv | AT aashishcgupta fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT guillaumecazoulat fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT maisaltaie fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT sireeshayedururi fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT bastienrigaud fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT austincastelo fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT johnwood fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT cenjiyu fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT caleboconnor fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT usamasalem fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT jessicaalbuquerquemarquessilva fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT aaronkylejones fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT mollymcculloch fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT brunocodisio fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT eugenejkoay fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages AT kristykbrock fullyautomateddeeplearningbasedautocontouringofliversegmentsandspleenoncontrastenhancedctimages |