Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.23805

Deep Learning Reconstruction for Accelerated Radial CEST MR Fingerprinting

Karandeep Cheema·Victor Murray·Ricardo Otazo·Ouri Cohen
View original source ↗Read source document ↗Export metadata ↓

Abstract

Purpose: To accelerate chemical exchange saturation transfer magnetic resonance fingerprinting (CEST-MRF) using temporal deep learning reconstruction. Methods: Ten healthy volunteers underwent repeated test-retest golden-angle radial CEST-MRF acquisitions with 30 measurements and 34 spokes per measurement. The Movienet network was adapted to reconstruct accelerated radial CEST-MRF data. Separate models were trained using 1, 2, 4, 8, 16, 20, and 34 spokes per measurement. Temporal and coil correlations were exploited by concatenating time and coil dimensions in the input. GRASP-based locally low-rank (GRASP-LLR) reconstruction was used for comparison. Quantitative DRONE fitting generated tissue parameter maps. Movienet-8 was evaluated with leave-one-subject-out analysis. Prospectively acquired 8-spoke datasets were compared with paired 34-spoke acquisitions, and Movienet-8 and Movienet-34 were compared with a 64-spoke acquisition. Results: In the held-out retrospective series, median Pearson correlations between Movienet and GRASP-LLR were 0.795, 0.947, 0.974, 0.977, and 0.981 for 1, 8, 16, 20, and 34 spokes, respectively. Corresponding NRMSE values were 14.7, 8.2, 5.9, 5.3, and 5.1 percent. Across 10 leave-one-subject-out volunteers, median image correlation was 0.965 and NRMSE was 7.5 percent. Reconstruction of a 30-slice series required 1.7 s with Movienet versus approximately 45 min with GRASP-LLR. Conclusion: Movienet enables whole-brain radial CEST-MRF acquisition in 3 min, with image reconstruction in less than 2 s.

Explore related subjects

physics.med-ph

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Karandeep Cheema, Victor Murray, Ricardo Otazo, Ouri Cohen. 2026-10-16. Deep Learning Reconstruction for Accelerated Radial CEST MR Fingerprinting. https://arxiv.org/abs/2610.23805

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Implementing the EIT complete electrode model through Robin boundary conditions

Objective. Medical electrical impedance tomography systems prescribe electrode currents. Accurate voltages, even on passive electrodes, require the complete electrode model (CEM), the most accurate model known for skin electrodes. Its current constraints specify total electrode currents, whereas standard finite element boundary interfaces impose local conditions. We describe how to implement the CEM through Robin boundary conditions, which are available in standard FEM packages. Approach. Prescribed electrode potentials give Robin problems whose unit-potential responses define an electrode transadmittance matrix. We prove equivalence to the conventional augmented CEM and formalise the discrete energy, equivalence and reference arguments in Lean 4. Numerical comparisons use two-dimensional P1 finite elements in DOLFINx, FreeFEM, GetFEM, MFEM and NGSolve, with the same nodal reference potential as EIDORS. Main results. For identical assembled blocks, the two procedures have the same solution set whenever the Robin matrix is invertible. Across the five packages in dimensionless test cases, maximum absolute electrode-potential differences from EIDORS were below $5\times10^{-13}$ (2692 nodes), and errors against an exact rational-arithmetic reference were below $1.1\times10^{-13}$ (2535 nodes). Significance. Standard Robin solves and consistent electrode-current integration provide a portable route to the current-driven CEM, with current constraints enforced through a small electrode system.

physics.med-ph↗
arXiv2026

Attenuation corrected activity reconstruction for listmode TOF-PET data

Time-of-flight (TOF) positron emission tomography (PET) emission data allow joint estimation of the activity and attenuation images. The methods proposed so far rely on reconstructions from TOF histograms of the measurements. However, with the rapid data increase of total-body (TB) scanners and the ever improving TOF and intrinsic resolutions, more memory and computation efficient techniques for joint reconstruction need to be developed. In this work, a modified version of the maximum likelihood activity and attenuation correction factors algorithm (MLACF) is presented, that jointly estimates the activity image together with the attenuation factors (AFs) from listmode emission data. The method does not require TOF histogramming of the data, however it still relies on a coarsely sampled (i.e. mashed) non-TOF histogram of the emission data. The activity distribution is subsequently quantified by reconstructing and enforcing known expected tissue attenuation on an attenuation reconstruction from the estimated AF values. The algorithm is evaluated using 3D simulations as well as clinical TOF-PET measurements of a patient scan and compared to reference reconstructions. As could be expected, we find the reconstruction accuracies improve with an "optimal" histogramming of the TOF emission data, increasing the signal to noise ratio of the data used for joint estimation. With an "optimal" histogramming of the measurements, the attenuation corrected activity distributions from TOF emission data only are comparable to the gold-standard in PET/CT.

physics.med-ph↗
arXiv2026

Localizing and quantifying cerebral arterial stenosis from sparse velocity waveforms using a conditional neural operator

Cerebral arterial stenosis alters blood flow throughout the vascular network, complicating the estimation of its location and severity from sparse velocity observations. We developed a conditional neural operator that predicts cerebral hemodynamics for candidate stenoses and enables inverse estimation through waveform comparison. Training data were generated using a sequential 0D-1D model that simulates pressure, cross-sectional area, and velocity under varying anatomical, physiological, and stenosis conditions. The operator encodes a single stenosis within vessel tokens and uses global attention to represent its effects across the network. Severity-gated coupling connects these representations to a shared field decoder, while paired healthy simulations constrain baseline predictions. After training, model weights remain fixed, and stenosis location and severity are estimated by minimizing the mismatch between predicted and observed midpoint velocity waveforms. Network-averaged relative validation errors were 0.14% for pressure, 0.59% for area, and 2.79% for velocity. For an eight-vessel observation configuration, localization accuracy across 20 candidate vessels was 99.4% without added noise and 96.0% at the 30% noise setting. Comparisons between configurations with equal numbers of observed vessels showed that measurement placement influenced recovery. Errors and localization ambiguities were greater in selected communicating and distal arteries. These results demonstrate the feasibility of surrogate-based stenosis estimation in simulated cerebral networks and highlight the importance of observation placement. Evaluation was limited to unseen stenosis scenarios within sampled anatomical and physiological conditions. Generalization to new individuals and clinical measurements requires further investigation.

physics.med-ph↗