Dataset Information

Three-dimensional self-attention conditional GAN with spectral normalization for multimodal neuroimaging synthesis.

ABSTRACT:

Purpose

To develop a new 3D generative adversarial network that is designed and optimized for the application of multimodal 3D neuroimaging synthesis.

Methods

We present a 3D conditional generative adversarial network (GAN) that uses spectral normalization and feature matching to stabilize the training process and ensure optimization convergence (called SC-GAN). A self-attention module was also added to model the relationships between widely separated image voxels. The performance of the network was evaluated on the data set from ADNI-3, in which the proposed network was used to predict PET images, fractional anisotropy, and mean diffusivity maps from multimodal MRI. Then, SC-GAN was applied on a multidimensional diffusion MRI experiment for superresolution application. Experiment results were evaluated by normalized RMS error, peak SNR, and structural similarity.

Results

In general, SC-GAN outperformed other state-of-the-art GAN networks including 3D conditional GAN in all three tasks across all evaluation metrics. Prediction error of the SC-GAN was 18%, 24% and 29% lower compared to 2D conditional GAN for fractional anisotropy, PET and mean diffusivity tasks, respectively. The ablation experiment showed that the major contributors to the improved performance of SC-GAN are the adversarial learning and the self-attention module, followed by the spectral normalization module. In the superresolution multidimensional diffusion experiment, SC-GAN provided superior predication in comparison to 3D Unet and 3D conditional GAN.

Conclusion

In this work, an efficient end-to-end framework for multimodal 3D medical image synthesis (SC-GAN) is presented. The source code is also made available at https://github.com/Haoyulance/SC-GAN.

SUBMITTER: Lan H

PROVIDER: S-EPMC9070032 | biostudies-literature | 2021 Sep

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Three-dimensional self-attention conditional GAN with spectral normalization for multimodal neuroimaging synthesis.

Lan Haoyu H Toga Arthur W AW Sepehrband Farshid F

Magnetic resonance in medicine 20210507 3

<h4>Purpose</h4>To develop a new 3D generative adversarial network that is designed and optimized for the application of multimodal 3D neuroimaging synthesis.<h4>Methods</h4>We present a 3D conditional generative adversarial network (GAN) that uses spectral normalization and feature matching to stabilize the training process and ensure optimization convergence (called SC-GAN). A self-attention module was also added to model the relationships between widely separated image voxels. The performance ...[more]

PMID: 33961321

Similar Datasets

Project description:ObjectivesLeveraging a large population-level morphologic, microstructural, and functional neuroimaging dataset, we aimed to elucidate the underlying neurobiology of attention-deficit hyperactivity disorder (ADHD) in children. In addition, we evaluated the applicability of machine learning classifiers to predict ADHD diagnosis based on imaging and clinical information.MethodsFrom the Adolescents Behavior Cognitive Development (ABCD) database, we included 1,798 children with ADHD diagnosis and 6,007 without ADHD. In multivariate logistic regression adjusted for age and sex, we examined the association of ADHD with different neuroimaging metrics. The neuroimaging metrics included fractional anisotropy (FA), neurite density (ND), mean-(MD), radial-(RD), and axial diffusivity (AD) of white matter (WM) tracts, cortical region thickness and surface areas from T1-MPRAGE series, and functional network connectivity correlations from resting-state fMRI.ResultsChildren with ADHD showed markers of pervasive reduced microstructural integrity in white matter (WM) with diminished neural density and fiber-tracks volumes - most notable in the frontal and parietal lobes. In addition, ADHD diagnosis was associated with reduced cortical volume and surface area, especially in the temporal and frontal regions. In functional MRI studies, ADHD children had reduced connectivity among default-mode network and the central and dorsal attention networks, which are implicated in concentration and attention function. The best performing combination of feature selection and machine learning classifier could achieve a receiver operating characteristics area under curve of 0.613 (95% confidence interval = 0.580-0.645) to predict ADHD diagnosis in independent validation, using a combination of multimodal imaging metrics and clinical variables.ConclusionOur study highlights the neurobiological implication of frontal lobe cortex and associate WM tracts in pathogenesis of childhood ADHD. We also demonstrated possible potentials and limitations of machine learning models to assist with ADHD diagnosis in a general population cohort based on multimodal neuroimaging metrics.

Project description:Study objectivesIn aging, reduced delta power (0.5-4 Hz) during N2 and N3 sleep has been associated with gray matter (GM) atrophy and hypometabolism within frontal regions. Some studies have also reported associations between N2 and N3 sleep delta power in specific sub-bands and amyloid pathology. Our objective was to better understand the relationships between spectral power in delta sub-bands during N2-N3 sleep and brain integrity using multimodal neuroimaging.MethodsIn-home polysomnography was performed in 127 cognitively unimpaired older adults (mean age ± SD: 69.0 ± 3.8 years). N2-N3 sleep EEG power was calculated in delta (0.5-4 Hz), slow delta (0.5-1 Hz), and fast delta (1-4 Hz) frequency bands. Participants also underwent magnetic resonance imaging and Florbetapir-PET (early and late acquisitions) scans to assess GM volume, brain perfusion, and amyloid burden. Amyloid accumulation over ~21 months was also quantified.ResultsHigher delta power was associated with higher GM volume mainly in fronto-cingular regions. Specifically, slow delta power was positively correlated with GM volume and perfusion in these regions, while the inverse association was observed with fast delta power. Delta power was neither associated with amyloid burden at baseline nor its accumulation over time, whatever the frequency band considered.ConclusionsOur results show that slow delta is particularly associated with preserved brain structure, and highlight the importance of analyzing delta power sub-bands to better understand the associations between delta power and brain integrity. Further longitudinal investigations with long follow-ups are needed to disentangle the associations among sleep, amyloid pathology, and dementia risk in older populations.Clinical trial informationName: Study in Cognitively Intact Seniors Aiming to Assess the Effects of Meditation Training (Age-Well). URL: https://clinicaltrials.gov/ct2/show/NCT02977819?term=Age-Well&draw=2&rank=1. See STROBE_statement_AGEWELL in supplemental materials.RegistrationEudraCT: 2016-002441-36; IDRCB: 2016-A01767-44; ClinicalTrials.gov Identifier: NCT02977819.

Dataset Information

Three-dimensional self-attention conditional GAN with spectral normalization for multimodal neuroimaging synthesis.

Purpose

Methods

Results

Conclusion

Publications

Three-dimensional self-attention conditional GAN with spectral normalization for multimodal neuroimaging synthesis.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets