SRNR: Training neural networks for Super-Resolution MRI using Noisy high-resolution Reference data

Jiaxin Xiao; Zihan Li; Berkin Bilgic; Jonathan R. Polimeni; Susie Huang; Qiyuan Tian

SRNR: Training neural networks for Super-Resolution MRI using Noisy high-resolution Reference data

Image and Video Processing 2022-11-11 v1

Authors: Jiaxin Xiao , Zihan Li , Berkin Bilgic , Jonathan R. Polimeni , Susie Huang , Qiyuan Tian

Abstract

Neural network (NN) based approaches for super-resolution MRI typically require high-SNR high-resolution reference data acquired in many subjects, which is time consuming and a barrier to feasible and accessible implementation. We propose to train NNs for Super-Resolution using Noisy Reference data (SRNR), leveraging the mechanism of the classic NN-based denoising method Noise2Noise. We systematically demonstrate that results from NNs trained using noisy and high-SNR references are similar for both simulated and empirical data. SRNR suggests a smaller number of repetitions of high-resolution reference data can be used to simplify the training data preparation for super-resolution MRI.

Keywords

image super-resolution image denoising magnetic resonance imaging reconstruction

Cite

@article{arxiv.2211.05360,
  title  = {SRNR: Training neural networks for Super-Resolution MRI using Noisy high-resolution Reference data},
  author = {Jiaxin Xiao and Zihan Li and Berkin Bilgic and Jonathan R. Polimeni and Susie Huang and Qiyuan Tian},
  journal= {arXiv preprint arXiv:2211.05360},
  year   = {2022}
}

Comments

2 pages, 5 figures, submitted to ISMRM

SRNR: Training neural networks for Super-Resolution MRI using Noisy high-resolution Reference data

Abstract

Keywords

Cite

Comments

Related papers