# ReconResNet: Regularised Residual Learning for MR Image Reconstruction of Undersampled Cartesian and Radial Data

arXiv: https://arxiv.org/abs/2103.09203 · HTML: https://mlsanity.com/paper/2103.09203/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 14 of its 58 references that are in the map.
- First posted on arXiv 16 Mar 2021.

## Builds on (most cited first, 14 of 14)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising](https://mlsanity.com/paper/1608.03981/) (2016)
- [Learning a Variational Network for Reconstruction of Accelerated MRI Data](https://mlsanity.com/paper/1704.00447/) (2017)
- [Efficient Object Localization Using Convolutional Networks](https://mlsanity.com/paper/1411.4280/) (2014)
- [Convolutional Recurrent Neural Networks for Dynamic MR Image Reconstruction](https://mlsanity.com/paper/1712.01751/) (2017)
- [Deep learning for undersampled MRI reconstruction](https://mlsanity.com/paper/1709.02576/) (2017)
- [Deep Residual Learning for Accelerated MRI using Magnitude and Phase Networks](https://mlsanity.com/paper/1804.00432/) (2018)
- [Learning-based Optimization of the Under-sampling Pattern in MRI](https://mlsanity.com/paper/1901.01960/) (2019)
- [DR2-Net: Deep Residual Reconstruction Network for Image Compressive Sensing](https://mlsanity.com/paper/1702.05743/) (2017)
- [Deep Residual Learning for Instrument Segmentation in Robotic Surgery](https://mlsanity.com/paper/1703.08580/) (2017)
- [Deep Generative Adversarial Networks for Compressed Sensing Automates MRI](https://mlsanity.com/paper/1706.00051/) (2017)
- [Cine Cardiac MRI Motion Artifact Reduction Using a Recurrent Neural Network](https://mlsanity.com/paper/2006.12700/) (2020)
- [OEDIPUS: An Experiment Design Framework for Sparsity-Constrained MRI](https://mlsanity.com/paper/1805.00524/) (2018)

## Built on by (most cited first, 4 of 4)

- [Sinogram upsampling using Primal-Dual UNet for undersampled CT and radial MRI reconstruction](https://mlsanity.com/paper/2112.13443/) (2021)
- [Weakly-supervised segmentation using inherently-explainable classification models and their application to brain tumour classification](https://mlsanity.com/paper/2206.05148/) (2022)
- [Complex Network for Complex Problems: A comparative study of CNN and Complex-valued CNN](https://mlsanity.com/paper/2302.04584/) (2023)
- [DDoS-UNet: Incorporating temporal information using Dynamic Dual-channel UNet for enhancing super-resolution of dynamic MRI](https://mlsanity.com/paper/2202.05355/) (2022)

## Abstract (arXiv)

MRI is an inherently slow process, which leads to long scan time for high-resolution imaging. The speed of acquisition can be increased by ignoring parts of the data (undersampling). Consequently, this leads to the degradation of image quality, such as loss of resolution or introduction of image artefacts. This work aims to reconstruct highly undersampled Cartesian or radial MR acquisitions, with better resolution and with less to no artefact compared to conventional techniques like compressed sensing. In recent times, deep learning has emerged as a very important area of research and has shown immense potential in solving inverse problems, e.g. MR image reconstruction. In this paper, a deep learning based MR image reconstruction framework is proposed, which includes a modified regularised version of ResNet as the network backbone to remove artefacts from the undersampled image, followed by data consistency steps that fusions the network output with the data already available from undersampled k-space in order to further improve reconstruction quality. The performance of this framework for various undersampling patterns has also been tested, and it has been observed that the framework is robust to deal with various sampling patterns, even when mixed together while training, and results in very high quality reconstruction, in terms of high SSIM (highest being 0.990±0.006 for acceleration factor of 3.5), while being compared with the fully sampled reconstruction. It has been shown that the proposed framework can successfully reconstruct even for an acceleration factor of 20 for Cartesian (0.968±0.005) and 17 for radially (0.962±0.012) sampled data. Furthermore, it has been shown that the framework preserves brain pathology during reconstruction while being trained on healthy subjects.