A deep learning algorithm to generate synthetic computed tomography images for brain treatments from 0.35 T magnetic resonance imaging
Background and Purpose: The development of Magnetic Resonance Imaging (MRI)-only Radiotherapy (RT) represents a significant advancement in the field. This study introduces a Deep Learning (DL) algorithm designed to quickly generate synthetic CT (sCT) images from low-field MR images in the brain, an...
Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-01-01
|
| Series: | Physics and Imaging in Radiation Oncology |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2405631625000132 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832576497419812864 |
|---|---|
| author | Luca Vellini Flaviovincenzo Quaranta Sebastiano Menna Elisa Pilloni Francesco Catucci Jacopo Lenkowicz Claudio Votta Michele Aquilano Andrea D’Aviero Martina Iezzi Francesco Preziosi Alessia Re Althea Boschetti Danila Piccari Antonio Piras Carmela Di Dio Alessandro Bombini Gian Carlo Mattiucci Davide Cusumano |
| author_facet | Luca Vellini Flaviovincenzo Quaranta Sebastiano Menna Elisa Pilloni Francesco Catucci Jacopo Lenkowicz Claudio Votta Michele Aquilano Andrea D’Aviero Martina Iezzi Francesco Preziosi Alessia Re Althea Boschetti Danila Piccari Antonio Piras Carmela Di Dio Alessandro Bombini Gian Carlo Mattiucci Davide Cusumano |
| author_sort | Luca Vellini |
| collection | DOAJ |
| description | Background and Purpose: The development of Magnetic Resonance Imaging (MRI)-only Radiotherapy (RT) represents a significant advancement in the field. This study introduces a Deep Learning (DL) algorithm designed to quickly generate synthetic CT (sCT) images from low-field MR images in the brain, an area not yet explored. Methods: Fifty-six patients were divided into training (32), validation (8), and test (16) groups. A conditional Generative Adversarial Network (cGAN) was trained on pre-processed axial paired images. sCTs were validated using mean absolute error (MAE) and mean error (ME) calculated within the patient body. Intensity Modulated Radiation Therapy (IMRT) plans were optimised on simulation MRI and calculated considering sCT and original CT as electron density (ED) map. Dose distributions using sCT and CT were compared using global gamma analysis at different tolerance criteria (2 %/2mm and 3 %/3mm) and evaluating the difference in estimating different Dose Volume Histogram (DVH) parameters for target and organs at risk (OARs). Results: The network generated sCTs of each single patient in less than two minutes (mean time = 103 ± 41 s). For test patients, the MAE was 62.1 ± 17.7 HU, and the ME was −7.3 ± 13.4 HU. Dose parameters on sCTs were within 0.5 Gy of those on original CTs. Gamma passing rates 2 %/2mm, and 3 %/3mm criteria were 99.5 %±0.5 %, and 99.7 %±0.3 %, respectively. Conclusion: The proposed DL algorithm generates in less than 2 min accurate sCT images in the brain for online adaptive radiotherapy, potentially eliminating the need for CT simulation in MR-only workflows for brain treatments. |
| format | Article |
| id | doaj-art-a2e490e7eaa2428f8a6d0364cde3c80d |
| institution | Kabale University |
| issn | 2405-6316 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Physics and Imaging in Radiation Oncology |
| spelling | doaj-art-a2e490e7eaa2428f8a6d0364cde3c80d2025-01-31T05:11:55ZengElsevierPhysics and Imaging in Radiation Oncology2405-63162025-01-0133100708A deep learning algorithm to generate synthetic computed tomography images for brain treatments from 0.35 T magnetic resonance imagingLuca Vellini0Flaviovincenzo Quaranta1Sebastiano Menna2Elisa Pilloni3Francesco Catucci4Jacopo Lenkowicz5Claudio Votta6Michele Aquilano7Andrea D’Aviero8Martina Iezzi9Francesco Preziosi10Alessia Re11Althea Boschetti12Danila Piccari13Antonio Piras14Carmela Di Dio15Alessandro Bombini16Gian Carlo Mattiucci17Davide Cusumano18Mater Olbia Hospital Olbia Sassari ItalyMater Olbia Hospital Olbia Sassari ItalyMater Olbia Hospital Olbia Sassari ItalyMater Olbia Hospital Olbia Sassari ItalyMater Olbia Hospital Olbia Sassari Italy; Corresponding author.Fondazione Policlinico Gemelli Agostino Gemelli IRCCS Roma ItalyFondazione Policlinico Gemelli Agostino Gemelli IRCCS Roma ItalyMater Olbia Hospital Olbia Sassari ItalyDepartment of Medical, Oral and Biotechnological Sciences, “Gabriele D’Annunzio” Università di Chieti, Italy; Department of Radiation Oncology, “S.S. Annunziata”, Chieti Hospital, ItalyMater Olbia Hospital Olbia Sassari ItalyMater Olbia Hospital Olbia Sassari ItalyMater Olbia Hospital Olbia Sassari ItalyMater Olbia Hospital Olbia Sassari ItalyMater Olbia Hospital Olbia Sassari ItalyUO Radioterapia Oncologica, Villa Santa Teresa Bagheria Palermo ItalyMater Olbia Hospital Olbia Sassari ItalyIstituto Nazionale di Fisica Nucleare (INFN) Sesto Fiorentino FI Italy; ICSC - Centro Nazionale di Ricerca in High Performance Computing, Big Data & Quantum Computing Casalecchio di Reno BO ItalyMater Olbia Hospital Olbia Sassari Italy; Università Cattolica del Sacro Cuore Rome ItalyMater Olbia Hospital Olbia Sassari ItalyBackground and Purpose: The development of Magnetic Resonance Imaging (MRI)-only Radiotherapy (RT) represents a significant advancement in the field. This study introduces a Deep Learning (DL) algorithm designed to quickly generate synthetic CT (sCT) images from low-field MR images in the brain, an area not yet explored. Methods: Fifty-six patients were divided into training (32), validation (8), and test (16) groups. A conditional Generative Adversarial Network (cGAN) was trained on pre-processed axial paired images. sCTs were validated using mean absolute error (MAE) and mean error (ME) calculated within the patient body. Intensity Modulated Radiation Therapy (IMRT) plans were optimised on simulation MRI and calculated considering sCT and original CT as electron density (ED) map. Dose distributions using sCT and CT were compared using global gamma analysis at different tolerance criteria (2 %/2mm and 3 %/3mm) and evaluating the difference in estimating different Dose Volume Histogram (DVH) parameters for target and organs at risk (OARs). Results: The network generated sCTs of each single patient in less than two minutes (mean time = 103 ± 41 s). For test patients, the MAE was 62.1 ± 17.7 HU, and the ME was −7.3 ± 13.4 HU. Dose parameters on sCTs were within 0.5 Gy of those on original CTs. Gamma passing rates 2 %/2mm, and 3 %/3mm criteria were 99.5 %±0.5 %, and 99.7 %±0.3 %, respectively. Conclusion: The proposed DL algorithm generates in less than 2 min accurate sCT images in the brain for online adaptive radiotherapy, potentially eliminating the need for CT simulation in MR-only workflows for brain treatments.http://www.sciencedirect.com/science/article/pii/S2405631625000132Artificial intelligenceSynthetic CTMRI-only workflowGAN |
| spellingShingle | Luca Vellini Flaviovincenzo Quaranta Sebastiano Menna Elisa Pilloni Francesco Catucci Jacopo Lenkowicz Claudio Votta Michele Aquilano Andrea D’Aviero Martina Iezzi Francesco Preziosi Alessia Re Althea Boschetti Danila Piccari Antonio Piras Carmela Di Dio Alessandro Bombini Gian Carlo Mattiucci Davide Cusumano A deep learning algorithm to generate synthetic computed tomography images for brain treatments from 0.35 T magnetic resonance imaging Physics and Imaging in Radiation Oncology Artificial intelligence Synthetic CT MRI-only workflow GAN |
| title | A deep learning algorithm to generate synthetic computed tomography images for brain treatments from 0.35 T magnetic resonance imaging |
| title_full | A deep learning algorithm to generate synthetic computed tomography images for brain treatments from 0.35 T magnetic resonance imaging |
| title_fullStr | A deep learning algorithm to generate synthetic computed tomography images for brain treatments from 0.35 T magnetic resonance imaging |
| title_full_unstemmed | A deep learning algorithm to generate synthetic computed tomography images for brain treatments from 0.35 T magnetic resonance imaging |
| title_short | A deep learning algorithm to generate synthetic computed tomography images for brain treatments from 0.35 T magnetic resonance imaging |
| title_sort | deep learning algorithm to generate synthetic computed tomography images for brain treatments from 0 35 t magnetic resonance imaging |
| topic | Artificial intelligence Synthetic CT MRI-only workflow GAN |
| url | http://www.sciencedirect.com/science/article/pii/S2405631625000132 |
| work_keys_str_mv | AT lucavellini adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT flaviovincenzoquaranta adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT sebastianomenna adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT elisapilloni adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT francescocatucci adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT jacopolenkowicz adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT claudiovotta adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT micheleaquilano adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT andreadaviero adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT martinaiezzi adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT francescopreziosi adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT alessiare adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT altheaboschetti adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT danilapiccari adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT antoniopiras adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT carmeladidio adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT alessandrobombini adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT giancarlomattiucci adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT davidecusumano adeeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT lucavellini deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT flaviovincenzoquaranta deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT sebastianomenna deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT elisapilloni deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT francescocatucci deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT jacopolenkowicz deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT claudiovotta deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT micheleaquilano deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT andreadaviero deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT martinaiezzi deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT francescopreziosi deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT alessiare deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT altheaboschetti deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT danilapiccari deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT antoniopiras deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT carmeladidio deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT alessandrobombini deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT giancarlomattiucci deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging AT davidecusumano deeplearningalgorithmtogeneratesyntheticcomputedtomographyimagesforbraintreatmentsfrom035tmagneticresonanceimaging |