ShuffleUNet: Super resolution of diffusion-weighted MRIs using deep learning

Chatterjee, Soumick; Sciarra, Alessandro; Dünnwald, Max; Mushunuri, Raghava Vinaykanth; Podishetti, Ranadheer; Rao, Rajatha Nagaraja; Doddapaneni, Gopinath, Geetha; Oeltze-Jafra, Steffen; Speck, Oliver; Nürnberger, Andreas

doi:10.23919/eusipco54536.2021.9615963

Cited by 20 publications

(23 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Originally proposed for image segmentation, different flavours of UNet have been developed and deployed in plenty of applications such as image segmentation (Milletari et al, 2016;Zhou et al, 2018;Oktay et al, 2018;Chatterjee et al, 2020a), audio source separation (Jansson et al, 2017;Stoller et al, 2018;Choi et al, 2019) and image reconstruction (Hyun et al, 2018;Iqbal et al, 2019). 3D UNet and its variants have been used for MR super-resolution as well (Pham et al, 2019;Sarasaen et al, 2021;Chatterjee et al, 2021b). Furthermore, UNet has been extended to multi-channel and dual-branch to incorporate prior-information (Chatterjee et al, 2020b).…”

Section: Related Workmentioning

confidence: 99%

DDoS-UNet: Incorporating temporal information using Dynamic Dual-channel UNet for enhancing super-resolution of dynamic MRI

Chatterjee¹,

Sarasaen²,

Rose³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Magnetic resonance imaging (MRI) provides high spatial resolution and excellent soft-tissue contrast without using harmful ionising radiation. Dynamic MRI is an essential tool for interventions to visualise movements or changes of the target organ. However, such MRI acquisition with high temporal resolution suffers from limited spatial resolution -also known as the spatio-temporal trade-off of dynamic MRI. Several approaches, including deep learning based super-resolution approaches, have been proposed to mitigate this trade-off. Nevertheless, such an approach typically aims to super-resolve each time-point separately, treating them as individual volumes. This research addresses the problem by creating a deep learning model which attempts to learn both spatial and temporal relationships. A modified 3D UNet model, DDoS-UNet, is proposed -which takes the low-resolution volume of the current time-point along with a prior image volume. Initially, the network is supplied with a static high-resolution planning scan as the prior image along with the low-resolution input to super-resolve the first time-point. Then it continues step-wise by using the super-resolved time-points as the prior image while super-resolving the subsequent time-points. The model performance was tested with 3D dynamic data that was undersampled to different in-plane levels. The proposed network achieved an average SSIM value of 0.951±0.017 while reconstructing the lowest resolution data (i.e. only 4% of the k-space acquired) -which could result in a theoretical acceleration factor of 25. The proposed approach can be used to reduce the required scan-time while achieving high spatial resolution -consequently alleviating the spatio-temporal trade-off of dynamic MRI, by incorporating prior knowledge of spatio-temporal information from the available high resolution planning scan and the existing temporal redundancy of time-series images into the network model.

show abstract

Section: Related Workmentioning

confidence: 99%

DDoS-UNet: Incorporating temporal information using Dynamic Dual-channel UNet for enhancing super-resolution of dynamic MRI

Chatterjee¹,

Sarasaen²,

Rose³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Their method is suitable for images with wide-range intensity. Chatterjee et al (2021) adopted UNet for training a super-resolution model and reconstructed the brain connectome with DWI results.…”

Section: Introductionmentioning

confidence: 99%

Improving the brain image resolution of generalized q-sampling MRI revealed by a three-dimensional CNN-based method

Shin

Chao²,

Li³

et al. 2023

Front. Neuroinform.

View full text Add to dashboard Cite

BackgroundUnderstanding neural connections facilitates the neuroscience and cognitive behavioral research. There are many nerve fiber intersections in the brain that need to be observed, and the size is between 30 and 50 nanometers. Improving image resolution has become an important issue for mapping the neural connections non-invasively. Generalized q-sampling imaging (GQI) was used to reveal the fiber geometry of straight and crossing. In this work, we attempted to achieve super-resolution with a deep learning method on diffusion weighted imaging (DWI).Materials and methodsA three-dimensional super-resolution convolutional neural network (3D SRCNN) was utilized to achieve super-resolution on DWI. Then, generalized fractional anisotropy (GFA), normalized quantitative anisotropy (NQA), and the isotropic value of the orientation distribution function (ISO) mapping were reconstructed using GQI with super-resolution DWI. We also reconstructed the orientation distribution function (ODF) of brain fibers using GQI.ResultsWith the proposed super-resolution method, the reconstructed DWI was closer to the target image than the interpolation method. The peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM) were also significantly improved. The diffusion index mapping reconstructed by GQI also had higher performance. The ventricles and white matter regions were much clearer.ConclusionThis super-resolution method can assist in postprocessing low-resolution images. With SRCNN, high-resolution images can be effectively and accurately generated. The method can clearly reconstruct the intersection structure in the brain connectome and has the potential to accurately describe the fiber geometry on a subvoxel scale.

show abstract

“…Over the last years, deep learning solutions have shown promising results for correcting motion in various MR applications. The MoCo problem has been approached as an image denoising problem, using convolutional neural networks as well as generative adversarial networks [2,12,14,20]. However, acting purely on image data, these methods cannot guarantee data consistency, which might hinder their translation into clinical practice.…”

Section: Introductionmentioning

confidence: 99%

Deep Learning-Based Detection of Motion-Affected k-Space Lines for T2*-Weighted MRI

Eichhorn¹,

Hammernik²,

Spieker³

et al. 2023

Preprint

View full text Add to dashboard Cite

T2*-weighted gradient echo MR imaging is strongly impacted by subject head motion due to motion-related changes in B0 inhomogeneities. Within the oxygenation-sensitive mqBOLD protocol, even mild motion during the acquisition of the T2*-weighted data propagates into errors in derived quantitative parameter maps. In order to correct these images without the need of repeated measurements, we propose to learn a classification of motion-affected k-space lines. To test this, we perform realistic motion simulations including motion-induced field inhomogeneity changes for supervised training. To detect the presence of motion in each phase encoding line, we train a convolutional neural network, leveraging the multi-echo information of the T2*-weighted images. The proposed network accurately detects motion-affected k-space lines for simulated displacements of ≥0.5 mm (accuracy on test set: 92.5 %). Finally, we show example reconstructions where we include these classification labels as weights in the data consistency term of an iterative reconstruction procedure, opening up exciting opportunities of k-space line detection in combination with more powerful reconstruction methods.

show abstract

ShuffleUNet: Super resolution of diffusion-weighted MRIs using deep learning

Cited by 20 publications

References 19 publications

DDoS-UNet: Incorporating temporal information using Dynamic Dual-channel UNet for enhancing super-resolution of dynamic MRI

DDoS-UNet: Incorporating temporal information using Dynamic Dual-channel UNet for enhancing super-resolution of dynamic MRI

Improving the brain image resolution of generalized q-sampling MRI revealed by a three-dimensional CNN-based method

Deep Learning-Based Detection of Motion-Affected k-Space Lines for T2*-Weighted MRI

Contact Info

Product

Resources

About