arXiv subjects

Baris Imre

Publications and source records attributed to Baris Imre.

6 recordsLinked to original sources

arXiv2026

Conditional Flow Matching for Cross-Field MRI Harmonisation

Magnetic resonance images of the same subject look markedly different across field strengths, which complicates the comparison and pooling of data across sites. We address cross-field brain-MRI translation for the MRIxFields2026 challenge, and in particular its Task~3: a single model that translates between any directed pair of the five field strengths and across three contrasts. We phrase the problem as a conditional flow matching path: because the source and target volumes are spatially registered, we learn a velocity field that carries the source slice directly to the target slice, rather than starting from noise. To learn this mapping from only three paired subjects, the unified model is trained in three stages: a degradation-bridge pretraining that distills a restoration prior from the abundant unpaired retrospective cohort, a cross-field finetuning over all directed pairs on the paired cohort, and an adversarial refinement that sharpens the output. At inference, we integrate the learned velocity with a second-order Heun solver in a handful of steps. A restoration prior learned without any paired data already reaches a mean SSIM of 0.837, and each subsequent training stage improves on it. A single 6.3M-parameter model thereby covers all 60 field-pair and contrast combinations, with inference in five solver steps per slice. On the challenge evaluation set the model reaches a mean SSIM of 0.909, averaged over the three contrasts, outperforming regression and diffusion baselines built on the identical network on all three challenge metrics.

cs.CV↗

arXiv2026

Deep Unrolled Networks in Representation Space Applied to MRI Reconstruction

Deep unrolled networks (DUNs) integrate physical forward models with learned regularization in cascaded network architectures, achieving exceptional performance in inverse problems while maintaining interpretability. While most DUNs operate in the object domain (e.g., image space), recent variants explored representation spaces for improved information flow. However, these methods rely on heuristic methods for data consistency (DC), sacrificing fidelity with measurements. In this work, we introduce DUNE (Deep Unrolled Networks in rEpresentation space), a framework that maintains exact adherence to physical measurements while operating in learned representation spaces. By deriving the DC gradient via the chain rule and implementing it through the Vector-Jacobian Product (VJP), we enable exact backpropagation of measurement residuals into the representation space. This formulation supports diverse architectural backbones, including pre-trained encoders to guide the iterative process. We assess DUNE against state-of-the-art baselines on accelerated MRI reconstruction tasks, demonstrating that exact VJP-based gradients yield superior reconstruction quality and structural fidelity across both single-channel portable low-field and multi-channel clinical high-field MRI acquisitions. The code will be available upon publication at https://github.com/EfeIlicak/DUNE.

eess.IV↗

arXiv2026

Physics-Guided Dual-Domain Network with Attention-Based Fusion for Portable MRI Reconstruction

Portable low-field magnetic resonance imaging (MRI) systems have gained renewed interest owing to their cost effectiveness and point-of-care imaging capabilities. Yet, portable MRI systems suffer from relatively low signal-to-noise ratio and limited hardware capabilities. While previous works have proposed the use of deep learning based reconstruction methods to improve low-field image quality, these operated only in the image-domain. Unlike other imaging modalities, MRI directly acquires data in the Fourier-domain (k-space), and exploiting both k-space and image-domain information can improve reconstruction quality. Here, we introduce DUN-DD, a novel physics-guided 3D network for portable MRI reconstruction, with parallel dual-domain branches whose outputs are combined together via an attention-based fusion network. To demonstrate the performance of the proposed method, we present \textit{in vivo} reconstructions obtained from both emulated datasets as well as images acquired with a 47mT Halbach-based portable MRI system. Our results show that DUN-DD outperforms state-of-the-art classical, data-driven, and physics-guided methods on both emulated and real portable MRI acquisitions.

physics.med-ph↗

arXiv2025

Physics-Informed Deep Unrolled Network for Portable MR Image Reconstruction

Magnetic resonance imaging (MRI) is the gold standard imaging modality for numerous diagnostic tasks, yet its usefulness is tempered due to its high cost and infrastructural requirements. Low-cost very-low-field portable scanners offer new opportunities, while enabling imaging outside conventional MRI suites. However, achieving diagnostic-quality images in clinically acceptable scan times remains challenging with these systems. Therefore methods for improving the image quality while reducing the scan duration are highly desirable. Here, we investigate a physics-informed 3D deep unrolled network for the reconstruction of portable MR acquisitions. Our approach includes a novel network architecture that utilizes momentum-based acceleration and leverages complex conjugate symmetry of k-space for improved reconstruction performance. Comprehensive evaluations on emulated datasets as well as 47mT portable MRI acquisitions demonstrate the improved reconstruction quality of the proposed method compared to existing methods.

physics.med-ph↗

arXiv2024

Brain-Shift: Unsupervised Pseudo-Healthy Brain Synthesis for Novel Biomarker Extraction in Chronic Subdural Hematoma

Chronic subdural hematoma (cSDH) is a common neurological condition characterized by the accumulation of blood between the brain and the dura mater. This accumulation of blood can exert pressure on the brain, potentially leading to fatal outcomes. Treatment options for cSDH are limited to invasive surgery or non-invasive management. Traditionally, the midline shift, hand-measured by experts from an ideal sagittal plane, and the hematoma volume have been the primary metrics for quantifying and analyzing cSDH. However, these approaches do not quantify the local 3D brain deformation caused by cSDH. We propose a novel method using anatomy-aware unsupervised diffeomorphic pseudo-healthy synthesis to generate brain deformation fields. The deformation fields derived from this process are utilized to extract biomarkers that quantify the shift in the brain due to cSDH. We use CT scans of 121 patients for training and validation of our method and find that our metrics allow the identification of patients who require surgery. Our results indicate that automatically obtained brain deformation fields might contain prognostic value for personalized cSDH treatment. Our implementation is available on: github.com/Barisimre/brain-morphing

eess.IV↗

arXiv2024

LaB-GATr: geometric algebra transformers for large biomedical surface and volume meshes

Many anatomical structures can be described by surface or volume meshes. Machine learning is a promising tool to extract information from these 3D models. However, high-fidelity meshes often contain hundreds of thousands of vertices, which creates unique challenges in building deep neural network architectures. Furthermore, patient-specific meshes may not be canonically aligned which limits the generalisation of machine learning algorithms. We propose LaB-GATr, a transfomer neural network with geometric tokenisation that can effectively learn with large-scale (bio-)medical surface and volume meshes through sequence compression and interpolation. Our method extends the recently proposed geometric algebra transformer (GATr) and thus respects all Euclidean symmetries, i.e. rotation, translation and reflection, effectively mitigating the problem of canonical alignment between patients. LaB-GATr achieves state-of-the-art results on three tasks in cardiovascular hemodynamics modelling and neurodevelopmental phenotype prediction, featuring meshes of up to 200,000 vertices. Our results demonstrate that LaB-GATr is a powerful architecture for learning with high-fidelity meshes which has the potential to enable interesting downstream applications. Our implementation is publicly available.

cs.CV↗