Dataset Information

Computed tomography synthesis from magnetic resonance imaging using cycle Generative Adversarial Networks with multicenter learning.

ABSTRACT: Background and Purpose: Addressing the need for accurate dose calculation in MRI-only radiotherapy, the generation of synthetic Computed Tomography (sCT) from MRI has emerged. Deep learning (DL) techniques, have shown promising results in achieving high sCT accuracies. However, existing sCT synthesis methods are often center-specific, posing a challenge to their generalizability. To overcome this limitation, recent studies have proposed approaches, such as multicenter training . Material and methods: The purpose of this work was to propose a multicenter sCT synthesis by DL, using a 2D cycle-GAN on 128 prostate cancer patients, from four different centers. Four cases were compared: monocenter cases, monocenter training and test on another center, multicenter trainings and a test on a center not included in the training and multicenter trainings with an included center in the test. Trainings were performed using 20 patients. sCT accuracy evaluation was performed using Mean Absolute Error, Mean Error and Peak-Signal-to-Noise-Ratio. Dose accuracy was assessed with gamma index and Dose Volume Histogram comparison. Results: Qualitative, quantitative and dose results show that the accuracy of sCTs for monocenter trainings and multicenter trainings using a seen center in the test did not differ significantly. However, when the test involved an unseen center, the sCT quality was inferior. Conclusions: The aim of this work was to propose generalizable multicenter training for MR-to-CT synthesis. It was shown that only a few data from one center included in the training cohort allows sCT accuracy equivalent to a monocenter study.

SUBMITTER: Texier B

PROVIDER: S-EPMC10709085 | biostudies-literature | 2023 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Computed tomography synthesis from magnetic resonance imaging using cycle Generative Adversarial Networks with multicenter learning.

Texier Blanche B Hémon Cédric C Lekieffre Pauline P Collot Emma E Tahri Safaa S Chourak Hilda H Dowling Jason J Greer Peter P Bessieres Igor I Acosta Oscar O Boue-Rafle Adrien A Guevelou Jennifer Le JL de Crevoisier Renaud R Lafond Caroline C Castelli Joël J Barateau Anaïs A Nunes Jean-Claude JC

Physics and imaging in radiation oncology 20231001

<b>Background and Purpose:</b> Addressing the need for accurate dose calculation in MRI-only radiotherapy, the generation of synthetic Computed Tomography (sCT) from MRI has emerged. Deep learning (DL) techniques, have shown promising results in achieving high sCT accuracies. However, existing sCT synthesis methods are often center-specific, posing a challenge to their generalizability. To overcome this limitation, recent studies have proposed approaches, such as multicenter training . <b>Materi ...[more]

PMID: 38077271

Similar Datasets

Project description:BackgroundLow-dose computed tomography (LDCT) is a diagnostic imaging technique designed to minimize radiation exposure to the patient. However, this reduction in radiation may compromise computed tomography (CT) image quality, adversely impacting clinical diagnoses. Various advanced LDCT methods have emerged to mitigate this challenge, relying on well-matched LDCT and normal-dose CT (NDCT) image pairs for training. Nevertheless, these methods often face difficulties in distinguishing image details from nonuniformly distributed noise, limiting their denoising efficacy. Additionally, acquiring suitably paired datasets in the medical domain poses challenges, further constraining their applicability. Hence, the objective of this study was to develop an innovative denoising framework for LDCT images employing unpaired data.MethodsIn this paper, we propose a LDCT denoising network (DNCNN) that alleviates the need for aligning LDCT and NDCT images. Our approach employs generative adversarial networks (GANs) to learn and model the noise present in LDCT images, establishing a mapping from the pseudo-LDCT to the actual NDCT domain without the need for paired CT images.ResultsWithin the domain of weakly supervised methods, our proposed model exhibited superior objective metrics on the simulated dataset when compared to CycleGAN and selective kernel-based cycle-consistent GAN (SKFCycleGAN): the peak signal-to-noise ratio (PSNR) was 43.9441, the structural similarity index measure (SSIM) was 0.9660, and the visual information fidelity (VIF) was 0.7707. In the clinical dataset, we conducted a visual effect analysis by observing various tissues through different observation windows. Our proposed method achieved a no-reference structural sharpness (NRSS) value of 0.6171, which was closest to that of the NDCT images (NRSS =0.6049), demonstrating its superiority over other denoising techniques in preserving details, maintaining structural integrity, and enhancing edge contrast.ConclusionsThrough extensive experiments on both simulated and clinical datasets, we demonstrated the superior efficacy of our proposed method in terms of denoising quality and quantity. Our method exhibits superiority over both supervised techniques, including block-matching and 3D filtering (BM3D), residual encoder-decoder convolutional neural network (RED-CNN), and Wasserstein generative adversarial network-VGG (WGAN-VGG), and over weakly supervised approaches, including CycleGAN and SKFCycleGAN.

Project description:BackgroundCardiovascular magnetic resonance (CMR) phase contrast is used to quantify blood flow. We sought to develop a complex-difference reconstruction for inline super-resolution of phase-contrast flow (CRISPFlow) to accelerate phase-contrast imaging.MethodsCRISPFlow was built on the super-resolution generative adversarial network. The model was trained and tested (4:1 ratio) using retrospectively identified phase-contrast images from 2020 patients (56 ± 16 years; 1131 men) referred for clinical 3T CMR at a single center from 2018 to 2023. For testing, ascending aortic flow images collected with 2.5 × 1.9 mm2 resolution using generalized autocalibrating partially parallel acquisitions (GRAPPA) were used to synthesize images with 7.5 × 1.9 mm2 resolution. CRISPFlow subsequently restored spatial resolution. In a prospective validation study of 38 participants (57 ± 15 years; 14 men) and 16 healthy individuals (42 ± 16 years; 6 men), CRISPFlow was applied to phase-contrast images collected with 7.5 × 1.9 mm2 resolution with use of GRAPPA and was compared to GRAPPA-accelerated images collected with 2.3 × 1.9 mm2 resolution. A blur metric was used to quantify sharpness. Aortic flow measurements were obtained semi-automatically. Statistical evaluation included analysis of variance, Bland-Altman analysis, and Pearson correlation coefficient (r).ResultsCRISPFlow reconstruction was successful in all cases. CRISPFlow reduced blurring in retrospective (0.35 vs 0.47, P < 0.001) and prospective (0.34 vs 0.48, P < 0.001) images with 7.5 × 1.9 mm2 resolution. Blurring in CRISPFlow images was similar to blurring in images with 2.5 × 1.9 mm2 (0.35 vs 0.35, P = 0.4082) and 2.3 × 1.9 mm2 (0.34 vs 0.32, P < 0.001) resolution. Bland-Altman differences in forward volume (-2 mL [-8 to 3 mL]), regurgitant volume (0 mL [-3 to 2 mL]), and a fraction (0% [-5 to 4%]) showed good agreement between the two techniques in a retrospective cohort. Differences in forward volume (1 mL [-11 to 14 ml]) also showed good agreement in the prospective cohort. There was a strong correlation (all r > 0.90) between GRAPPA and CRISPFlow measurements of flow in both studies.ConclusionWe demonstrated the potential of CRISPFlow to accelerate phase contrast CMR.

Dataset Information

Computed tomography synthesis from magnetic resonance imaging using cycle Generative Adversarial Networks with multicenter learning.

Publications

Computed tomography synthesis from magnetic resonance imaging using cycle Generative Adversarial Networks with multicenter learning.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets