Loading
Life sciences · Preprint
arXiv · September 6, 2026
Early or partial results. Treat as a signal, not a conclusion.
This preprint describes Shared LoRA, a parameter-efficient deep learning framework for accelerated MRI reconstruction that uses a single shared adapter set and gating network to handle multiple acceleration factors simultaneously. The method achieves competitive image quality metrics (PSNR, SSIM) across factors with only 5.3% of model parameters, but clinical utility, generalization to real-world acquisition protocols, and peer-reviewed validation are not yet established.
Computational method validation study. Accelerated MRI k-space data; undersampled inputs generated by random sampling of acceleration factors and corresponding sampling masks during training.. Intervention: Shared LoRA: frozen SHFormer backbone with shared LoRA adapters and lightweight gating network (GateNet) that generates layer-wise coefficients to modulate residual strength..
Shared LoRA achieves best or competitive PSNR and SSIM across acceleration factors Trainable parameters account for only 5.3% of total model parameters Performance at lower acceleration factors remains largely unaffected as jointly trained factor set expands
Safety was not reported in the material analysed. Check the source before drawing any conclusion about harm.
If validated in peer review and clinical studies, this parameter-efficient approach could reduce computational cost and storage for multi-factor MRI reconstruction systems. Clinical relevance depends on whether PSNR/SSIM improvements translate to diagnostic accuracy gains.
This is a first-report technical validation of a novel machine learning architecture for MRI reconstruction, demonstrating feasibility and comparative metrics but lacking clinical outcome validation or peer review.
As stated by the source record.
Quoted from the source exactly as published.
If validated in peer review and clinical studies, this parameter-efficient approach could reduce computational cost and storage for multi-factor MRI reconstruction systems. Clinical relevance depends on whether PSNR/SSIM improvements translate to diagnostic accuracy gains.
Graded across the dimensions that decide whether you should act, each from what the source actually supports. There is no single score, and where a dimension was not assessed it says so.
Accelerated MRI reconstruction recovers images from undersampled k-space. However, different acceleration factors produce distinct artifact patterns. Existing methods often train separate models for each factor, leading to poor cross-factor generalization and high training and storage costs. We propose Shared LoRA, a parameter-efficient framework that freezes the pretrained SHFormer backbone and trains a single shared set of LoRA adapters together with a lightweight gating network. During training, undersampled inputs are generated by randomly sampling acceleration factors and their corresponding sampling masks, enabling the shared adapters to learn reconstruction knowledge across factors. Given the acceleration factor, GateNet generates layer-wise coefficients to dynamically modulate the residual strength of each adapter. Experiments show that Shared LoRA achieves the best or competitive PSNR and SSIM across acceleration factors, while its trainable parameters account for only about 5.3% of the total model parameters. Its performance at lower acceleration factors remains largely unaffected as the jointly trained factor set expands, and it generalizes stably to unseen neighboring factors.
Taken from the source record, never inferred. Follow any of these and new work involving them reaches your briefing.