Compressed Sensing MRI Reconstruction with Co-VeGAN: Complex-Valued Generative Adversarial Network

Vasudeva, Bhavya; Deora, Puneesh; Bhattacharya, Sumanta; Pradhan, Pyari Mohan

doi:10.1109/wacv51458.2022.00184

Cited by 17 publications

(10 citation statements)

References 39 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They also comment on complex differentiability, referring to earlier works [18]. Complex-valued building blocks have been used to develop a multitude of architectures, such as complex-valued generative adversarial networks [2,3], complexvalued convolutional recurrent networks [19] and a complex-valued U-net [4]. There has also been recent interest in optimizing computability on GPUs for complex-valued neural networks [20].…”

Section: Related Workmentioning

confidence: 99%

“…Most of these methods focus on real-valued pipelines for applications with real-valued signals, such as natural images or encodings of natural language processing. There is however a great amount of applications that naturally deal with complex-valued signals, such as MRI images [1,2] or remote sensing [3] and the Fourier transform of real-valued signals [4,5] or images [6,7] and it has been shown that fully complex-valued architectures often (but not always [8]) deliver superior performance when dealing with complex-valued signals. The complex numbers come with an intrinsic algebraic structure that can not be captured by the simple isomorphism of C ∼ R 2 , especially because there is no natural way to define multiplication in R 2 , which, however, is an important part of many deep learning building blocks.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Building Blocks for a Complex-Valued Transformer Architecture

Florian¹,

Jiang²

2023

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

Most deep learning pipelines are built on real-valued operations to deal with real-valued inputs such as images, speech or music signals. However, a lot of applications naturally make use of complex-valued signals or images, such as MRI or remote sensing. Additionally the Fourier transform of signals is complex-valued and has numerous applications. We aim to make deep learning directly applicable to these complex-valued signals without using projections into R 2 . Thus we add to the recent developments of complex-valued neural networks by presenting building blocks to transfer the transformer architecture to the complex domain. We present multiple versions of a complex-valued Scaled Dot-Product Attention mechanism as well as a complex-valued layer normalization. We test on a classification and a sequence generation task on the MusicNet dataset and show improved robustness to overfitting while maintaining on-par performance when compared to the real-valued transformer architecture.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Building Blocks for a Complex-Valued Transformer Architecture

Florian¹,

Jiang²

2023

ICASSP 2023 - 2023 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)

View full text Add to dashboard Cite

show abstract

“…For example, the projection-based cascaded U-Net model [ 23 ]; the geometric distillation network, used to unfold the model-based CS-MRI optimization problem [ 24 ]; the iterative fusion model, used to integrate the image- and gradient-based priors into reconstruction [ 25 ]; the interpretable network, which has two-grid cycle and geometric prior distillation [ 26 ]; and the fast iterative shrinkage thresholding network, used for high throughput reconstruction [ 27 ]. The new GAN-powered algorithms have also continued to grow [ 29 , 30 , 31 , 32 ], including ESSGAN [ 29 ], DBGAN [ 30 ], CoVeGAN [ 31 ], and SEPGAN [ 32 ]. In ESSGAN [ 29 ], structurally strengthened connections were introduced to enhance feature propagation and reuse the in-between concatenated convolutional autoencoders in addition to residual blocks.…”

Section: Introductionmentioning

confidence: 99%

“…In DBGAN [ 30 ], the dual-branch GAN model uses cross-stage skip connection between two end-to-end-cascaded U-Nets to widen the channels for feature propagation in the generator. In the complex-valued GAN (Co-VeGAN) network [ 31 ], the use of complex-valued weights and operations was explored in addition to the use of a complex-valued activation function that is sensitive to the input phase. In SepGAN [ 32 ], depth-wise separable convolution was utilized as the basic component to reduce the number of learning parameters.…”

Section: Introductionmentioning

confidence: 99%

CS-MRI Reconstruction Using an Improved GAN with Dilated Residual Networks and Channel Attention Mechanism

Li,

Zhang,

Yang

et al. 2023

Sensors

View full text Add to dashboard Cite

Compressed sensing (CS) MRI has shown great potential in enhancing time efficiency. Deep learning techniques, specifically generative adversarial networks (GANs), have emerged as potent tools for speedy CS-MRI reconstruction. Yet, as the complexity of deep learning reconstruction models increases, this can lead to prolonged reconstruction time and challenges in achieving convergence. In this study, we present a novel GAN-based model that delivers superior performance without the model complexity escalating. Our generator module, built on the U-net architecture, incorporates dilated residual (DR) networks, thus expanding the network’s receptive field without increasing parameters or computational load. At every step of the downsampling path, this revamped generator module includes a DR network, with the dilation rates adjusted according to the depth of the network layer. Moreover, we have introduced a channel attention mechanism (CAM) to distinguish between channels and reduce background noise, thereby focusing on key information. This mechanism adeptly combines global maximum and average pooling approaches to refine channel attention. We conducted comprehensive experiments with the designed model using public domain MRI datasets of the human brain. Ablation studies affirmed the efficacy of the modified modules within the network. Incorporating DR networks and CAM elevated the peak signal-to-noise ratios (PSNR) of the reconstructed images by about 1.2 and 0.8 dB, respectively, on average, even at 10× CS acceleration. Compared to other relevant models, our proposed model exhibits exceptional performance, achieving not only excellent stability but also outperforming most of the compared networks in terms of PSNR and SSIM. When compared with U-net, DR-CAM-GAN’s average gains in SSIM and PSNR were 14% and 15%, respectively. Its MSE was reduced by a factor that ranged from two to seven. The model presents a promising pathway for enhancing the efficiency and quality of CS-MRI reconstruction.

show abstract

“…On the other hand, in [ 7 ], X-ray chest images were denoised using complex-valued neural networks, showing the high capacity of this kind of structure in health applications. In [ 15 , 16 , 17 , 18 ], the authors used the CVDL and complex-valued data to detect brain diseases using fMRI data, reaching outstanding performance. The key differentiating factor of this work compared with our proposed one is the nature of the input data, because we are using skin images and scalograms built from heart sounds.…”

Section: Introductionmentioning

confidence: 99%

A Fair Performance Comparison between Complex-Valued and Real-Valued Neural Networks for Disease Detection

2022

View full text Add to dashboard Cite

Our aim is to contribute to the classification of anomalous patterns in biosignals using this novel approach. We specifically focus on melanoma and heart murmurs. We use a comparative study of two convolution networks in the Complex and Real numerical domains. The idea is to obtain a powerful approach for building portable systems for early disease detection. Two similar algorithmic structures were chosen so that there is no bias determined by the number of parameters to train. Three clinical data sets, ISIC2017, PH2, and Pascal, were used to carry out the experiments. Mean comparison hypothesis tests were performed to ensure statistical objectivity in the conclusions. In all cases, complex-valued networks presented a superior performance for the Precision, Recall, F1 Score, Accuracy, and Specificity metrics in the detection of associated anomalies. The best complex number-based classifier obtained in the Receiving Operating Characteristic (ROC) space presents a Euclidean distance of 0.26127 with respect to the ideal classifier, as opposed to the best real number-based classifier, whose Euclidean distance to the ideal is 0.36022 for the same task of melanoma detection. The 27.46% superiority in this metric, as in the others reported in this work, suggests that complex-valued networks have a greater ability to extract features for more efficient discrimination in the dataset.

show abstract

Compressed Sensing MRI Reconstruction with Co-VeGAN: Complex-Valued Generative Adversarial Network

Cited by 17 publications

References 39 publications

Building Blocks for a Complex-Valued Transformer Architecture

Building Blocks for a Complex-Valued Transformer Architecture

CS-MRI Reconstruction Using an Improved GAN with Dilated Residual Networks and Channel Attention Mechanism

A Fair Performance Comparison between Complex-Valued and Real-Valued Neural Networks for Disease Detection

Contact Info

Product

Resources

About