A deep learning model to generate synthetic CT for prostate MR-only radiotherapy dose planning: a multicenter study
IntroductionFor radiotherapy based solely on magnetic resonance imaging (MRI), generating synthetic computed tomography scans (sCT) from MRI is essential for dose calculation. The use of deep learning (DL) methods to generate sCT from MRI has shown encouraging results if the MRI images used for trai...
| Main Authors: | , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2023-11-01
|
| Series: | Frontiers in Oncology |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fonc.2023.1279750/full |
| _version_ | 1827631583326109696 |
|---|---|
| author | Safaa Tahri Blanche Texier Jean-Claude Nunes Cédric Hemon Pauline Lekieffre Emma Collot Hilda Chourak Hilda Chourak Jennifer Le Guevelou Peter Greer Peter Greer Jason Dowling Oscar Acosta Igor Bessieres Louis Marage Adrien Boue-Rafle Renaud De Crevoisier Caroline Lafond Anaïs Barateau |
| author_facet | Safaa Tahri Blanche Texier Jean-Claude Nunes Cédric Hemon Pauline Lekieffre Emma Collot Hilda Chourak Hilda Chourak Jennifer Le Guevelou Peter Greer Peter Greer Jason Dowling Oscar Acosta Igor Bessieres Louis Marage Adrien Boue-Rafle Renaud De Crevoisier Caroline Lafond Anaïs Barateau |
| author_sort | Safaa Tahri |
| collection | DOAJ |
| description | IntroductionFor radiotherapy based solely on magnetic resonance imaging (MRI), generating synthetic computed tomography scans (sCT) from MRI is essential for dose calculation. The use of deep learning (DL) methods to generate sCT from MRI has shown encouraging results if the MRI images used for training the deep learning network and the MRI images for sCT generation come from the same MRI device. The objective of this study was to create and evaluate a generic DL model capable of generating sCTs from various MRI devices for prostate radiotherapyMaterials and methodsIn total, 90 patients from three centers (30 CT-MR prostate pairs/center) underwent treatment using volumetric modulated arc therapy for prostate cancer (PCa) (60 Gy in 20 fractions). T2 MRI images were acquired in addition to computed tomography (CT) images for treatment planning. The DL model was a 2D supervised conditional generative adversarial network (Pix2Pix). Patient images underwent preprocessing steps, including nonrigid registration. Seven different supervised models were trained, incorporating patients from one, two, or three centers. Each model was trained on 24 CT-MR prostate pairs. A generic model was trained using patients from all three centers. To compare sCT and CT, the mean absolute error in Hounsfield units was calculated for the entire pelvis, prostate, bladder, rectum, and bones. For dose analysis, mean dose differences of D99% for CTV, V95% for PTV, Dmax for rectum and bladder, and 3D gamma analysis (local, 1%/1 mm) were calculated from CT and sCT. Furthermore, Wilcoxon tests were performed to compare the image and dose results obtained with the generic model to those with the other trained models.ResultsConsidering the image results for the entire pelvis, when the data used for the test comes from the same center as the data used for training, the results were not significantly different from the generic model. Absolute dose differences were less than 1 Gy for the CTV D99% for every trained model and center. The gamma analysis results showed nonsignificant differences between the generic and monocentric models.ConclusionThe accuracy of sCT, in terms of image and dose, is equivalent to whether MRI images are generated using the generic model or the monocentric model. The generic model, using only eight MRI-CT pairs per center, offers robust sCT generation, facilitating PCa MRI-only radiotherapy for routine clinical use. |
| first_indexed | 2024-03-09T14:23:49Z |
| format | Article |
| id | doaj.art-a1d51ded1429495bb38f6ed943c64561 |
| institution | Directory Open Access Journal |
| issn | 2234-943X |
| language | English |
| last_indexed | 2024-03-09T14:23:49Z |
| publishDate | 2023-11-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Oncology |
| spelling | doaj.art-a1d51ded1429495bb38f6ed943c645612023-11-28T08:32:17ZengFrontiers Media S.A.Frontiers in Oncology2234-943X2023-11-011310.3389/fonc.2023.12797501279750A deep learning model to generate synthetic CT for prostate MR-only radiotherapy dose planning: a multicenter studySafaa Tahri0Blanche Texier1Jean-Claude Nunes2Cédric Hemon3Pauline Lekieffre4Emma Collot5Hilda Chourak6Hilda Chourak7Jennifer Le Guevelou8Peter Greer9Peter Greer10Jason Dowling11Oscar Acosta12Igor Bessieres13Louis Marage14Adrien Boue-Rafle15Renaud De Crevoisier16Caroline Lafond17Anaïs Barateau18University of Rennes, Centre de Lutte contre le Cancer (CLCC) Eugène Marquis, INSERM Laboratoire Traitement du Signal et de l'Image (LTSI) - Unité Mixte de Recherche (UMR) 1099, Rennes, FranceUniversity of Rennes, Centre de Lutte contre le Cancer (CLCC) Eugène Marquis, INSERM Laboratoire Traitement du Signal et de l'Image (LTSI) - Unité Mixte de Recherche (UMR) 1099, Rennes, FranceUniversity of Rennes, Centre de Lutte contre le Cancer (CLCC) Eugène Marquis, INSERM Laboratoire Traitement du Signal et de l'Image (LTSI) - Unité Mixte de Recherche (UMR) 1099, Rennes, FranceUniversity of Rennes, Centre de Lutte contre le Cancer (CLCC) Eugène Marquis, INSERM Laboratoire Traitement du Signal et de l'Image (LTSI) - Unité Mixte de Recherche (UMR) 1099, Rennes, FranceUniversity of Rennes, Centre de Lutte contre le Cancer (CLCC) Eugène Marquis, INSERM Laboratoire Traitement du Signal et de l'Image (LTSI) - Unité Mixte de Recherche (UMR) 1099, Rennes, FranceUniversity of Rennes, Centre de Lutte contre le Cancer (CLCC) Eugène Marquis, INSERM Laboratoire Traitement du Signal et de l'Image (LTSI) - Unité Mixte de Recherche (UMR) 1099, Rennes, FranceUniversity of Rennes, Centre de Lutte contre le Cancer (CLCC) Eugène Marquis, INSERM Laboratoire Traitement du Signal et de l'Image (LTSI) - Unité Mixte de Recherche (UMR) 1099, Rennes, FranceThe Australian eHealth Research Centre, Commonwealth Scientific and Industrial Research Organisation (CSIRO), Health and Biosecurity, Brisbane, QLD, AustraliaUniversity of Rennes, Centre de Lutte contre le Cancer (CLCC) Eugène Marquis, INSERM Laboratoire Traitement du Signal et de l'Image (LTSI) - Unité Mixte de Recherche (UMR) 1099, Rennes, FranceSchool of Mathematical and Physical Sciences, University of Newcastle, Newcastle, NSW, AustraliaRadiation Oncology, Calvary Mater Newcastle Hospital, Newcastle, NSW, AustraliaThe Australian eHealth Research Centre, Commonwealth Scientific and Industrial Research Organisation (CSIRO), Health and Biosecurity, Brisbane, QLD, AustraliaUniversity of Rennes, Centre de Lutte contre le Cancer (CLCC) Eugène Marquis, INSERM Laboratoire Traitement du Signal et de l'Image (LTSI) - Unité Mixte de Recherche (UMR) 1099, Rennes, FranceCentre Georges François Leclerc, Dijon, FranceCentre Georges François Leclerc, Dijon, FranceUniversity of Rennes, Centre de Lutte contre le Cancer (CLCC) Eugène Marquis, INSERM Laboratoire Traitement du Signal et de l'Image (LTSI) - Unité Mixte de Recherche (UMR) 1099, Rennes, FranceUniversity of Rennes, Centre de Lutte contre le Cancer (CLCC) Eugène Marquis, INSERM Laboratoire Traitement du Signal et de l'Image (LTSI) - Unité Mixte de Recherche (UMR) 1099, Rennes, FranceUniversity of Rennes, Centre de Lutte contre le Cancer (CLCC) Eugène Marquis, INSERM Laboratoire Traitement du Signal et de l'Image (LTSI) - Unité Mixte de Recherche (UMR) 1099, Rennes, FranceUniversity of Rennes, Centre de Lutte contre le Cancer (CLCC) Eugène Marquis, INSERM Laboratoire Traitement du Signal et de l'Image (LTSI) - Unité Mixte de Recherche (UMR) 1099, Rennes, FranceIntroductionFor radiotherapy based solely on magnetic resonance imaging (MRI), generating synthetic computed tomography scans (sCT) from MRI is essential for dose calculation. The use of deep learning (DL) methods to generate sCT from MRI has shown encouraging results if the MRI images used for training the deep learning network and the MRI images for sCT generation come from the same MRI device. The objective of this study was to create and evaluate a generic DL model capable of generating sCTs from various MRI devices for prostate radiotherapyMaterials and methodsIn total, 90 patients from three centers (30 CT-MR prostate pairs/center) underwent treatment using volumetric modulated arc therapy for prostate cancer (PCa) (60 Gy in 20 fractions). T2 MRI images were acquired in addition to computed tomography (CT) images for treatment planning. The DL model was a 2D supervised conditional generative adversarial network (Pix2Pix). Patient images underwent preprocessing steps, including nonrigid registration. Seven different supervised models were trained, incorporating patients from one, two, or three centers. Each model was trained on 24 CT-MR prostate pairs. A generic model was trained using patients from all three centers. To compare sCT and CT, the mean absolute error in Hounsfield units was calculated for the entire pelvis, prostate, bladder, rectum, and bones. For dose analysis, mean dose differences of D99% for CTV, V95% for PTV, Dmax for rectum and bladder, and 3D gamma analysis (local, 1%/1 mm) were calculated from CT and sCT. Furthermore, Wilcoxon tests were performed to compare the image and dose results obtained with the generic model to those with the other trained models.ResultsConsidering the image results for the entire pelvis, when the data used for the test comes from the same center as the data used for training, the results were not significantly different from the generic model. Absolute dose differences were less than 1 Gy for the CTV D99% for every trained model and center. The gamma analysis results showed nonsignificant differences between the generic and monocentric models.ConclusionThe accuracy of sCT, in terms of image and dose, is equivalent to whether MRI images are generated using the generic model or the monocentric model. The generic model, using only eight MRI-CT pairs per center, offers robust sCT generation, facilitating PCa MRI-only radiotherapy for routine clinical use.https://www.frontiersin.org/articles/10.3389/fonc.2023.1279750/fullMR-only radiotherapydose planningMRIdeep learningCT synthesis |
| spellingShingle | Safaa Tahri Blanche Texier Jean-Claude Nunes Cédric Hemon Pauline Lekieffre Emma Collot Hilda Chourak Hilda Chourak Jennifer Le Guevelou Peter Greer Peter Greer Jason Dowling Oscar Acosta Igor Bessieres Louis Marage Adrien Boue-Rafle Renaud De Crevoisier Caroline Lafond Anaïs Barateau A deep learning model to generate synthetic CT for prostate MR-only radiotherapy dose planning: a multicenter study Frontiers in Oncology MR-only radiotherapy dose planning MRI deep learning CT synthesis |
| title | A deep learning model to generate synthetic CT for prostate MR-only radiotherapy dose planning: a multicenter study |
| title_full | A deep learning model to generate synthetic CT for prostate MR-only radiotherapy dose planning: a multicenter study |
| title_fullStr | A deep learning model to generate synthetic CT for prostate MR-only radiotherapy dose planning: a multicenter study |
| title_full_unstemmed | A deep learning model to generate synthetic CT for prostate MR-only radiotherapy dose planning: a multicenter study |
| title_short | A deep learning model to generate synthetic CT for prostate MR-only radiotherapy dose planning: a multicenter study |
| title_sort | deep learning model to generate synthetic ct for prostate mr only radiotherapy dose planning a multicenter study |
| topic | MR-only radiotherapy dose planning MRI deep learning CT synthesis |
| url | https://www.frontiersin.org/articles/10.3389/fonc.2023.1279750/full |
| work_keys_str_mv | AT safaatahri adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT blanchetexier adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT jeanclaudenunes adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT cedrichemon adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT paulinelekieffre adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT emmacollot adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT hildachourak adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT hildachourak adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT jenniferleguevelou adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT petergreer adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT petergreer adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT jasondowling adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT oscaracosta adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT igorbessieres adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT louismarage adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT adrienbouerafle adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT renauddecrevoisier adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT carolinelafond adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT anaisbarateau adeeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT safaatahri deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT blanchetexier deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT jeanclaudenunes deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT cedrichemon deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT paulinelekieffre deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT emmacollot deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT hildachourak deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT hildachourak deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT jenniferleguevelou deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT petergreer deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT petergreer deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT jasondowling deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT oscaracosta deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT igorbessieres deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT louismarage deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT adrienbouerafle deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT renauddecrevoisier deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT carolinelafond deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy AT anaisbarateau deeplearningmodeltogeneratesyntheticctforprostatemronlyradiotherapydoseplanningamulticenterstudy |