-
Generative Autoregressive Transformers for Model-Agnostic Federated MRI Reconstruction
Authors:
Valiyeh A. Nezhad,
Gokberk Elmas,
Bilal Kabas,
Fuat Arslan,
Tolga Çukur
Abstract:
Although learning-based models hold great promise for MRI reconstruction, single-site models built on limited local datasets often suffer from poor generalization. This challenge has spurred interest in collaborative model training on multi-site datasets via federated learning (FL) -- a privacy-preserving framework that aggregates model updates instead of sharing imaging data. Conventional FL aggr…
▽ More
Although learning-based models hold great promise for MRI reconstruction, single-site models built on limited local datasets often suffer from poor generalization. This challenge has spurred interest in collaborative model training on multi-site datasets via federated learning (FL) -- a privacy-preserving framework that aggregates model updates instead of sharing imaging data. Conventional FL aggregates locally trained model weights into a global model, inherently constraining all sites to use a homogeneous model architecture. This rigidity forces sites to compromise on architectures tailored to their compute resources and application-specific needs, making conventional FL unsuitable for model-heterogeneous settings where each site may prefer a distinct architecture. To overcome this limitation, we introduce FedGAT, a novel model-agnostic FL technique based on generative autoregressive transformers. FedGAT decentralizes the training of a global generative prior that learns the distribution of multi-site MR images. For high-fidelity synthesis, we propose a novel site-prompted GAT prior that controllably synthesizes realistic MR images from desired sites via autoregressive prediction across spatial scales. Each site then trains its own reconstruction model -- using an architecture of its choice -- on a hybrid dataset augmenting its local MRI dataset with GAT-generated synthetic MR images emulating datasets from other sites. This hybrid training strategy enables site-specific reconstruction models to generalize more effectively across diverse data distributions while preserving data privacy. Comprehensive experiments on multi-institutional datasets demonstrate that FedGAT enables flexible, model-heterogeneous collaborations and achieves superior within-site and cross-site reconstruction performance compared to state-of-the-art FL baselines.
△ Less
Submitted 8 May, 2025; v1 submitted 6 February, 2025;
originally announced February 2025.
-
Learning Fourier-Constrained Diffusion Bridges for MRI Reconstruction
Authors:
Muhammad U. Mirza,
Onat Dalmaz,
Hasan A. Bedel,
Gokberk Elmas,
Yilmaz Korkmaz,
Alper Gungor,
Salman UH Dar,
Tolga Çukur
Abstract:
Deep generative models have gained recent traction in accelerated MRI reconstruction. Diffusion priors are particularly promising given their representational fidelity. Instead of the target transformation from undersampled to fully-sampled data required for MRI reconstruction, common diffusion priors are trained to learn a task-agnostic transformation from an asymptotic start-point of Gaussian no…
▽ More
Deep generative models have gained recent traction in accelerated MRI reconstruction. Diffusion priors are particularly promising given their representational fidelity. Instead of the target transformation from undersampled to fully-sampled data required for MRI reconstruction, common diffusion priors are trained to learn a task-agnostic transformation from an asymptotic start-point of Gaussian noise onto the finite end-point of fully-sampled data. During inference, data-consistency projections are injected in between reverse diffusion steps to reach a compromise solution within the span of both the trained diffusion prior and the imaging operator for an accelerated MRI acquisition. Unfortunately, performance losses can occur due to the discrepancy between target and learned transformations given the asymptotic normality assumption in diffusion priors. To address this discrepancy, here we introduce a novel Fourier-constrained diffusion bridge (FDB) for MRI reconstruction that transforms between a finite start-point of moderately undersampled data and an end-point of fully-sampled data. We derive the theoretical formulation of FDB as a generalized diffusion process based on a stochastic degradation operator that performs random spatial-frequency removal. We propose an enhanced sampling algorithm with a learned correction term for soft dealiasing across reverse diffusion steps. Demonstrations on brain MRI indicate that FDB outperforms state-of-the-art methods including non-diffusion and diffusion priors.
△ Less
Submitted 16 December, 2023; v1 submitted 2 August, 2023;
originally announced August 2023.
-
One Model to Unite Them All: Personalized Federated Learning of Multi-Contrast MRI Synthesis
Authors:
Onat Dalmaz,
Usama Mirza,
Gökberk Elmas,
Muzaffer Özbey,
Salman UH Dar,
Emir Ceyani,
Salman Avestimehr,
Tolga Çukur
Abstract:
Multi-institutional collaborations are key for learning generalizable MRI synthesis models that translate source- onto target-contrast images. To facilitate collaboration, federated learning (FL) adopts decentralized training and mitigates privacy concerns by avoiding sharing of imaging data. However, FL-trained synthesis models can be impaired by the inherent heterogeneity in the data distributio…
▽ More
Multi-institutional collaborations are key for learning generalizable MRI synthesis models that translate source- onto target-contrast images. To facilitate collaboration, federated learning (FL) adopts decentralized training and mitigates privacy concerns by avoiding sharing of imaging data. However, FL-trained synthesis models can be impaired by the inherent heterogeneity in the data distribution, with domain shifts evident when common or variable translation tasks are prescribed across sites. Here we introduce the first personalized FL method for MRI Synthesis (pFLSynth) to improve reliability against domain shifts. pFLSynth is based on an adversarial model that produces latents specific to individual sites and source-target contrasts, and leverages novel personalization blocks to adaptively tune the statistics and weighting of feature maps across the generator stages given latents. To further promote site specificity, partial model aggregation is employed over downstream layers of the generator while upstream layers are retained locally. As such, pFLSynth enables training of a unified synthesis model that can reliably generalize across multiple sites and translation tasks. Comprehensive experiments on multi-site datasets clearly demonstrate the enhanced performance of pFLSynth against prior federated methods in multi-contrast MRI synthesis.
△ Less
Submitted 23 August, 2022; v1 submitted 13 July, 2022;
originally announced July 2022.
-
Adaptive Diffusion Priors for Accelerated MRI Reconstruction
Authors:
Alper Güngör,
Salman UH Dar,
Şaban Öztürk,
Yilmaz Korkmaz,
Gokberk Elmas,
Muzaffer Özbey,
Tolga Çukur
Abstract:
Deep MRI reconstruction is commonly performed with conditional models that de-alias undersampled acquisitions to recover images consistent with fully-sampled data. Since conditional models are trained with knowledge of the imaging operator, they can show poor generalization across variable operators. Unconditional models instead learn generative image priors decoupled from the operator to improve…
▽ More
Deep MRI reconstruction is commonly performed with conditional models that de-alias undersampled acquisitions to recover images consistent with fully-sampled data. Since conditional models are trained with knowledge of the imaging operator, they can show poor generalization across variable operators. Unconditional models instead learn generative image priors decoupled from the operator to improve reliability against domain shifts related to the imaging operator. Recent diffusion models are particularly promising given their high sample fidelity. Nevertheless, inference with a static image prior can perform suboptimally. Here we propose the first adaptive diffusion prior for MRI reconstruction, AdaDiff, to improve performance and reliability against domain shifts. AdaDiff leverages an efficient diffusion prior trained via adversarial mapping over large reverse diffusion steps. A two-phase reconstruction is executed following training: a rapid-diffusion phase that produces an initial reconstruction with the trained prior, and an adaptation phase that further refines the result by updating the prior to minimize data-consistency loss. Demonstrations on multi-contrast brain MRI clearly indicate that AdaDiff outperforms competing conditional and unconditional methods under domain shifts, and achieves superior or on par within-domain performance.
△ Less
Submitted 17 September, 2023; v1 submitted 12 July, 2022;
originally announced July 2022.
-
Federated Learning of Generative Image Priors for MRI Reconstruction
Authors:
Gokberk Elmas,
Salman UH Dar,
Yilmaz Korkmaz,
Emir Ceyani,
Burak Susam,
Muzaffer Özbey,
Salman Avestimehr,
Tolga Çukur
Abstract:
Multi-institutional efforts can facilitate training of deep MRI reconstruction models, albeit privacy risks arise during cross-site sharing of imaging data. Federated learning (FL) has recently been introduced to address privacy concerns by enabling distributed training without transfer of imaging data. Existing FL methods for MRI reconstruction employ conditional models to map from undersampled t…
▽ More
Multi-institutional efforts can facilitate training of deep MRI reconstruction models, albeit privacy risks arise during cross-site sharing of imaging data. Federated learning (FL) has recently been introduced to address privacy concerns by enabling distributed training without transfer of imaging data. Existing FL methods for MRI reconstruction employ conditional models to map from undersampled to fully-sampled acquisitions via explicit knowledge of the imaging operator. Since conditional models generalize poorly across different acceleration rates or sampling densities, imaging operators must be fixed between training and testing, and they are typically matched across sites. To improve generalization and flexibility in multi-institutional collaborations, here we introduce a novel method for MRI reconstruction based on Federated learning of Generative IMage Priors (FedGIMP). FedGIMP leverages a two-stage approach: cross-site learning of a generative MRI prior, and subject-specific injection of the imaging operator. The global MRI prior is learned via an unconditional adversarial model that synthesizes high-quality MR images based on latent variables. Specificity in the prior is preserved via a mapper subnetwork that produces site-specific latents. During inference, the prior is combined with subject-specific imaging operators to enable reconstruction, and further adapted to individual test samples by minimizing data-consistency loss. Comprehensive experiments on multi-institutional datasets clearly demonstrate enhanced generalization performance of FedGIMP against site-specific and federated methods based on conditional models, as well as traditional reconstruction methods.
△ Less
Submitted 6 April, 2022; v1 submitted 8 February, 2022;
originally announced February 2022.
