Transfer learning enhanced generative adversarial networks for multi-channel MRI reconstruction

The state-of-the-art in cardiac MRI reconstruction: Results of the CMRxRecon challenge in MICCAI 2023

Cardiac magnetic resonance imaging (MRI) provides detailed and quantitative evaluation of the heart's structure, function, and tissue characteristics with high-resolution spatial-temporal imaging. However, its slow imaging speed and motion artifacts are notable limitations. Undersampling reconstruction, especially data-driven algorithms, has emerged as a promising solution to accelerate scans and enhance imaging performance using highly under-sampled data. Nevertheless, the scarcity of publicly available cardiac k-space datasets and evaluation platform hinder the development of data-driven reconstruction algorithms. To address this issue, we organized the Cardiac MRI Reconstruction Challenge (CMRxRecon) in 2023, in collaboration with the 26th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI). CMRxRecon presented an extensive k-space dataset comprising cine and mapping raw data, accompanied by detailed annotations of cardiac anatomical structures. With overwhelming participation, the challenge attracted more than 285 teams and over 600 participants. Among them, 22 teams successfully submitted Docker containers for the testing phase, with 7 teams submitted for both cine and mapping tasks. All teams use deep learning based approaches, indicating that deep learning has predominately become a promising solution for the problem. The first-place winner of both tasks utilizes the E2E-VarNet architecture as backbones. In contrast, U-Net is still the most popular backbone for both multi-coil and single-coil reconstructions. This paper provides a comprehensive overview of the challenge design, presents a summary of the submitted results, reviews the employed methods, and offers an in-depth discussion that aims to inspire future advancements in cardiac MRI reconstruction models. The summary emphasizes the effective strategies observed in Cardiac MRI reconstruction, including backbone architecture, loss function, pre-processing techniques, physical modeling, and model complexity, thereby providing valuable insights for further developments in this field.

Cardiac MR image reconstruction using cascaded hybrid dual domain deep learning framework

Recovering diagnostic-quality cardiac MR images from highly under-sampled data is a current research focus, particularly in addressing cardiac and respiratory motion. Techniques such as Compressed Sensing (CS) and Parallel Imaging (pMRI) have been proposed to accelerate MRI data acquisition and improve image quality. However, these methods have limitations in high spatial-resolution applications, often resulting in blurring or residual artifacts. Recently, deep learning-based techniques have gained attention for their accuracy and efficiency in image reconstruction. Deep learning-based MR image reconstruction methods are divided into two categories: (a) single domain methods (image domain learning and k-space domain learning) and (b) cross/dual domain methods. Single domain methods, which typically use U-Net in either the image or k-space domain, fail to fully exploit the correlation between these domains. This paper introduces a dual-domain deep learning approach that incorporates multi-coil data consistency (MCDC) layers for reconstructing cardiac MR images from 1-D Variable Density (VD) random under-sampled data. The proposed hybrid dual-domain deep learning models integrate data from both the domains to improve image quality, reduce artifacts, and enhance overall robustness and accuracy of the reconstruction process. Experimental results demonstrate that the proposed methods outperform than conventional deep learning and CS techniques, as evidenced by higher Structural Similarity Index (SSIM), lower Root Mean Square Error (RMSE), and higher Peak Signal-to-Noise Ratio (PSNR).

A review of deep learning-based reconstruction methods for accelerated MRI using spatiotemporal and multi-contrast redundancies

Accelerated magnetic resonance imaging (MRI) has played an essential role in reducing data acquisition time for MRI. Acceleration can be achieved by acquiring fewer data points in k-space, which results in various artifacts in the image domain. Conventional reconstruction methods have resolved the artifacts by utilizing multi-coil information, but with limited robustness. Recently, numerous deep learning-based reconstruction methods have been developed, enabling outstanding reconstruction performances with higher acceleration. Advances in hardware and developments of specialized network architectures have produced such achievements. Besides, MRI signals contain various redundant information including multi-coil redundancy, multi-contrast redundancy, and spatiotemporal redundancy. Utilization of the redundant information combined with deep learning approaches allow not only higher acceleration, but also well-preserved details in the reconstructed images. Consequently, this review paper introduces the basic concepts of deep learning and conventional accelerated MRI reconstruction methods, followed by review of recent deep learning-based reconstruction methods that exploit various redundancies. Lastly, the paper concludes by discussing the challenges, limitations, and potential directions of future developments.

Knowledge-driven deep learning for fast MR imaging: Undersampled MR image reconstruction from supervised to un-supervised learning

Deep learning (DL) has emerged as a leading approach in accelerating MRI. It employs deep neural networks to extract knowledge from available datasets and then applies the trained networks to reconstruct accurate images from limited measurements. Unlike natural image restoration problems, MRI involves physics-based imaging processes, unique data properties, and diverse imaging tasks. This domain knowledge needs to be integrated with data-driven approaches. Our review will introduce the significant challenges faced by such knowledge-driven DL approaches in the context of fast MRI along with several notable solutions, which include learning neural networks and addressing different imaging application scenarios. The traits and trends of these techniques have also been given which have shifted from supervised learning to semi-supervised learning, and finally, to unsupervised learning methods. In addition, MR vendors' choices of DL reconstruction have been provided along with some discussions on open questions and future directions, which are critical for the reliable imaging systems.

Transfer Learning in Cancer Genetics, Mutation Detection, Gene Expression Analysis, and Syndrome Recognition

Artificial intelligence (AI), encompassing machine learning (ML) and deep learning (DL), has revolutionized medical research, facilitating advancements in drug discovery and cancer diagnosis. ML identifies patterns in data, while DL employs neural networks for intricate processing. Predictive modeling challenges, such as data labeling, are addressed by transfer learning (TL), leveraging pre-existing models for faster training. TL shows potential in genetic research, improving tasks like gene expression analysis, mutation detection, genetic syndrome recognition, and genotype-phenotype association. This review explores the role of TL in overcoming challenges in mutation detection, genetic syndrome detection, gene expression, or phenotype-genotype association. TL has shown effectiveness in various aspects of genetic research. TL enhances the accuracy and efficiency of mutation detection, aiding in the identification of genetic abnormalities. TL can improve the diagnostic accuracy of syndrome-related genetic patterns. Moreover, TL plays a crucial role in gene expression analysis in order to accurately predict gene expression levels and their interactions. Additionally, TL enhances phenotype-genotype association studies by leveraging pre-trained models. In conclusion, TL enhances AI efficiency by improving mutation prediction, gene expression analysis, and genetic syndrome detection. Future studies should focus on increasing domain similarities, expanding databases, and incorporating clinical data for better predictions.

MA-RECON: Mask-aware deep-neural-network for robust fast MRI k-space interpolation

BACKGROUND AND OBJECTIVE

High-quality reconstruction of MRI images from under-sampled 'k-space' data, which is in the Fourier domain, is crucial for shortening MRI acquisition times and ensuring superior temporal resolution. Over recent years, a wealth of deep neural network (DNN) methods have emerged, aiming to tackle the complex, ill-posed inverse problem linked to this process. However, their instability against variations in the acquisition process and anatomical distribution exposes a deficiency in the generalization of relevant physical models within these DNN architectures. The goal of our work is to enhance the generalization capabilities of DNN methods for k-space interpolation by introducing 'MA-RECON', an innovative mask-aware DNN architecture and associated training method.

METHODS

Unlike preceding approaches, our 'MA-RECON' architecture encodes not only the observed data but also the under-sampling mask within the model structure. It implements a tailored training approach that leverages data generated with a variety of under-sampling masks to stimulate the model's generalization of the under-sampled MRI reconstruction problem. Therefore, effectively represents the associated inverse problem, akin to the classical compressed sensing approach.

RESULTS

The benefits of our MA-RECON approach were affirmed through rigorous testing with the widely accessible fastMRI dataset. Compared to standard DNN methods and DNNs trained with under-sampling mask augmentation, our approach demonstrated superior generalization capabilities. This resulted in a considerable improvement in robustness against variations in both the acquisition process and anatomical distribution, especially in regions with pathology.

CONCLUSION

In conclusion, our mask-aware strategy holds promise for enhancing the generalization capacity and robustness of DNN-based methodologies for MRI reconstruction from undersampled k-space data. Code is available in the following link: https://github.com/nitzanavidan/PD_Recon.

Kernel extreme learning with harmonized bat algorithm for prediction of pyrene toxicity in rats

Polycyclic aromatic hydrocarbons (PAHs) are organic pollutants and manufactured substances conferring toxicity to human health. The present study investigated whether pyrene, a type of PAH, harms rats. Our research provides an effective feature selection strategy for the animal dataset from Wenzhou Medical University's Experimental Animal Center to thoroughly examine the impacts of PAH toxicity on rat features. Initially, we devised a high-performance optimization method (SCBA) and added the Sobol sequence, vertical crossover and horizontal crossover mechanisms to the bat algorithm (BA). The SCBA-KELM model, which combines SCBA with the kernel extreme learning machine model (KELM), has excellent accuracy and high stability for selecting features. Benchmark function tests are then used in this research to verify the overall optimization performance of SCBA. In this paper, the feature selection performance of SCBA-KELM is verified using various comparative experiments. According to the results, the features of the genes PXR, CAR, CYP2B1/2 and CYP1A1/2 have the most impact on rats. The SCBA-KELM model's classification performance for the gene dataset was 100%, and the model's precision value for the public dataset was around 96%, as determined by the classification index. In conclusion, the model utilized in this research is anticipated to be a reliable and valuable approach for toxicological classification and assessment.

A Systematic Review and Identification of the Challenges of Deep Learning Techniques for Undersampled Magnetic Resonance Image Reconstruction

Deep learning (DL) in magnetic resonance imaging (MRI) shows excellent performance in image reconstruction from undersampled k-space data. Artifact-free and high-quality MRI reconstruction is essential for ensuring accurate diagnosis, supporting clinical decision-making, enhancing patient safety, facilitating efficient workflows, and contributing to the validity of research studies and clinical trials. Recently, deep learning has demonstrated several advantages over conventional MRI reconstruction methods. Conventional methods rely on manual feature engineering to capture complex patterns and are usually computationally demanding due to their iterative nature. Conversely, DL methods use neural networks with hundreds of thousands of parameters and automatically learn relevant features and representations directly from the data. Nevertheless, there are some limitations to DL-based techniques concerning MRI reconstruction tasks, such as the need for large, labeled datasets, the possibility of overfitting, and the complexity of model training. Researchers are striving to develop DL models that are more efficient, adaptable, and capable of providing valuable information for medical practitioners. We provide a comprehensive overview of the current developments and clinical uses by focusing on state-of-the-art DL architectures and tools used in MRI reconstruction. This study has three objectives. Our main objective is to describe how various DL designs have changed over time and talk about cutting-edge tactics, including their advantages and disadvantages. Hence, data pre- and post-processing approaches are assessed using publicly available MRI datasets and source codes. Secondly, this work aims to provide an extensive overview of the ongoing research on transformers and deep convolutional neural networks for rapid MRI reconstruction. Thirdly, we discuss several network training strategies, like supervised, unsupervised, transfer learning, and federated learning for rapid and efficient MRI reconstruction. Consequently, this article provides significant resources for future improvement of MRI data pre-processing and fast image reconstruction.

Face Image Segmentation Using Boosted Grey Wolf Optimizer

Image segmentation methods have received widespread attention in face image recognition, which can divide each pixel in the image into different regions and effectively distinguish the face region from the background for further recognition. Threshold segmentation, a common image segmentation method, suffers from the problem that the computational complexity shows exponential growth with the increase in the segmentation threshold level. Therefore, in order to improve the segmentation quality and obtain the segmentation thresholds more efficiently, a multi-threshold image segmentation framework based on a meta-heuristic optimization technique combined with Kapur's entropy is proposed in this study. A meta-heuristic optimization method based on an improved grey wolf optimizer variant is proposed to optimize the 2D Kapur's entropy of the greyscale and nonlocal mean 2D histograms generated by image computation. In order to verify the advancement of the method, experiments compared with the state-of-the-art method on IEEE CEC2020 and face image segmentation public dataset were conducted in this paper. The proposed method has achieved better results than other methods in various tests at 18 thresholds with an average feature similarity of 0.8792, an average structural similarity of 0.8532, and an average peak signal-to-noise ratio of 24.9 dB. It can be used as an effective tool for face segmentation.

Effects of radiofrequency channel numbers on B1+ mapping using the Bloch-Siegert shift method

PURPOSE

This paper aims to investigate the impact of the channel numbers on the performance of B1+ mapping, by using the Bloch-Siegert shift (BSS) method. B1+ mapping plays a crucial role in various brain imaging protocols.

THEORY AND METHODS

We simulated the radiofrequency field of the human head model in six groups of multi-channel receive coil with a range of different channel numbers. MR signals were synthesized according to the standard BSS sequence, with quantified Gaussian added. Next, we combined the signals of each channel to reconstruct the B1+ map by weighted averaging and maximum likelihood estimation strategies and evaluate the bias by relative standard deviation of each coil.

RESULTS

The simulation results revealed that the accuracy of B1+ maps improved with the increasing of channel numbers, meanwhile the per channel efficiency of B1+maps accuracy gradually decrease. Both trends slowed down when the channel numbers reached 12 or above.

CONCLUSION

Our finding suggests that increasing the channel numbers can improve the accuracy of B1+map. However, a diminishing efficiency of per channel accuracy improvement was overserved, indicating that the relationship between quality of B1+ map and the channel numbers is nonlinear. Based on these findings, our study provides a reference for determining channel numbers to achieve a balance of coil selection and manufacturing cost. It also provides a theoretical basis for evaluating other B1+ mapping techniques.

Adaptive channel-modulated personalized federated learning for magnetic resonance image reconstruction

Magnetic resonance imaging (MRI) is extensively utilized in clinical practice for diagnostic purposes, owing to its non-invasive nature and remarkable ability to provide detailed characterization of soft tissues. However, its drawback lies in the prolonged scanning time. To accelerate MR imaging, how to reconstruct MR images from under-sampled data quickly and accurately has drawn intensive research interest; it, however, remains a challenging task. While some deep learning models have achieved promising performance in MRI reconstruction, these models usually require a substantial quantity of paired data for training, which proves challenging to gather and share owing to high scanning costs and data privacy concerns. Federated learning (FL) is a potential tool to alleviate these difficulties. It enables multiple clinical clients to collaboratively train a global model without compromising privacy. However, it is extremely challenging to fit a single model to diverse data distributions of different clients. Moreover, existing FL algorithms treat the features of each channel equally, lacking discriminative learning ability across feature channels, and hence hindering their representational capability. In this study, we propose a novel Adaptive Channel-Modulated Federal learning framework for personalized MRI reconstruction, dubbed as ACM-FedMRI. Specifically, considering each local client may focus on features in different channels, we first design a client-specific hypernetwork to guide the channel selection operation in order to optimize the extracted features. Additionally, we introduce a performance-based channel decoupling scheme, which dynamically separates the global model at the channel level to facilitate personalized adjustments based on the performance of individual clients. This approach eliminates the need for heuristic design of specific personalization layers. Extensive experiments on four datasets under two different settings show that our ACM-FedMRI achieves outstanding results compared to other cutting-edge federated learning techniques in the field of MRI reconstruction.

Anti-sine-cosine atom search optimization (ASCASO): a novel approach for parameter estimation of PV models

Nowadays, solar power generation has gradually become a part of electric energy sharing. How to effectively enhance the energy conversion efficiency of solar cells and components has gradually emerged as a focal point of research. This paper presents a boosted atomic search optimization (ASO) with a new anti-sine-cosine mechanism (ASCASO) to realize the parameter estimation of photovoltaic (PV) models. The anti-sine-cosine mechanism is inspired by the update principle of sine cosine algorithm (SCA) and the mutation strategy of linear population size reduction adaptive differential evolution (LSHADE). The working principle of anti-sine-cosine mechanism is to utilize two mutation formulas containing arcsine and arccosine functions to further update the position of atoms. The introduction of anti-sine-cosine mechanism achieves the populations' random handover and promotes the neighbors' information communication. For better evaluation, the proposed ASCASO is devoted to estimate parameters of three PV models of R.T.C France, one Photowat-PWP201 PV module model, and two commercial polycrystalline PV panels including STM6-40/36 and STM6-120/36 with monocrystalline cells. The proposed ASCASO is compared with nine reported comparative algorithms to assess the performance. The results of parameter estimation for different PV models of various methods demonstrate that ASCASO performs more accurately and reliably than other reported comparative methods. Thus, ASCASO can be considered a highly effective approach for accurately estimating the parameters of PV models.

An enhanced grey wolf optimizer boosted machine learning prediction model for patient-flow prediction

Large and medium-sized general hospitals have adopted artificial intelligence big data systems to optimize the management of medical resources to improve the quality of hospital outpatient services and decrease patient wait times in recent years as a result of the development of medical information technology and the rise of big medical data. However, owing to the impact of several elements, including the physical environment, patient, and physician behaviours, the real optimum treatment effect does not meet expectations. In order to promote orderly patient access, this work provides a patient-flow prediction model that takes into account shifting dynamics and objective rules of patient-flow to handle this issue and forecast patients' medical requirements. First, we propose a high-performance optimization method (SRXGWO) and integrate the Sobol sequence, Cauchy random replacement strategy, and directional mutation mechanism into the grey wolf optimization (GWO) algorithm. The patient-flow prediction model (SRXGWO-SVR) is then proposed using SRXGWO to optimize the parameters of support vector regression (SVR). Twelve high-performance algorithms are examined in the benchmark function experiments' ablation and peer algorithm comparison tests, which are intended to validate SRXGWO's optimization performance. In order to forecast independently in the patient-flow prediction trials, the data set is split into training and test sets. The findings demonstrated that SRXGWO-SVR outperformed the other seven peer models in terms of prediction accuracy and error. As a result, SRXGWO-SVR is anticipated to be a reliable and efficient patient-flow forecast system that may help hospitals manage medical resources as effectively as possible.

Prediction of fractional flow reserve with enhanced ant lion optimized support vector machine

The evaluation of coronary morphology provides important guidance for the treatment of coronary heart disease (CHD). A chaotic Gaussian mutation antlion optimizer algorithm (CGALO) is proposed in the paper, and it is combined with SVM to construct a classification prediction model for Fractional flow reserve (FFR). To overcome the limitations of the original antlion optimizer (ALO) algorithm, the chaotic Gaussian mutation strategy is introduced, which leads to an improvement in its convergence speed and accuracy. To evaluate the proposed algorithm's performance, comparative experiments were conducted on 23 benchmark functions alongside 12 other cutting-edge optimization algorithms. The experimental outcomes demonstrate that the proposed algorithm achieves superior convergence accuracy and speed compared to the alternative comparison algorithms. Additionally, it is combined with SVM and FS to construct a hierarchical FFR classification model, which is utilized to make effective predictions for 84 patients at the affiliated hospital of medical school, Ningbo university. The experimental results demonstrate that the proposed model achieves an average accuracy of 92%. Moreover, it concludes that smoking history, number of lesion vessels, lesion location, diffuse lesions and ST segment changes, and other factors are the most critical indicators for FFR. Therefore, the model that has been established is a new FFR intelligent classification prediction technology that can effectively assist doctors in making corresponding decisions and evaluation plans.

Improved Colony Predation Algorithm Optimized Convolutional Neural Networks for Electrocardiogram Signal Classification

Recently, swarm intelligence algorithms have received much attention because of their flexibility for solving complex problems in the real world. Recently, a new algorithm called the colony predation algorithm (CPA) has been proposed, taking inspiration from the predatory habits of groups in nature. However, CPA suffers from poor exploratory ability and cannot always escape solutions known as local optima. Therefore, to improve the global search capability of CPA, an improved variant (OLCPA) incorporating an orthogonal learning strategy is proposed in this paper. Then, considering the fact that the swarm intelligence algorithm can go beyond the local optimum and find the global optimum solution, a novel OLCPA-CNN model is proposed, which uses the OLCPA algorithm to tune the parameters of the convolutional neural network. To verify the performance of OLCPA, comparison experiments are designed to compare with other traditional metaheuristics and advanced algorithms on IEEE CEC 2017 benchmark functions. The experimental results show that OLCPA ranks first in performance compared to the other algorithms. Additionally, the OLCPA-CNN model achieves high accuracy rates of 97.7% and 97.8% in classifying the MIT-BIH Arrhythmia and European ST-T datasets.

Linear fine-tuning: a linear transformation based transfer strategy for deep MRI reconstruction

Introduction

Fine-tuning (FT) is a generally adopted transfer learning method for deep learning-based magnetic resonance imaging (MRI) reconstruction. In this approach, the reconstruction model is initialized with pre-trained weights derived from a source domain with ample data and subsequently updated with limited data from the target domain. However, the direct full-weight update strategy can pose the risk of "catastrophic forgetting" and overfitting, hindering its effectiveness. The goal of this study is to develop a zero-weight update transfer strategy to preserve pre-trained generic knowledge and reduce overfitting.

Methods

Based on the commonality between the source and target domains, we assume a linear transformation relationship of the optimal model weights from the source domain to the target domain. Accordingly, we propose a novel transfer strategy, linear fine-tuning (LFT), which introduces scaling and shifting (SS) factors into the pre-trained model. In contrast to FT, LFT only updates SS factors in the transfer phase, while the pre-trained weights remain fixed.

Results

To evaluate the proposed LFT, we designed three different transfer scenarios and conducted a comparative analysis of FT, LFT, and other methods at various sampling rates and data volumes. In the transfer scenario between different contrasts, LFT outperforms typical transfer strategies at various sampling rates and considerably reduces artifacts on reconstructed images. In transfer scenarios between different slice directions or anatomical structures, LFT surpasses the FT method, particularly when the target domain contains a decreasing number of training images, with a maximum improvement of up to 2.06 dB (5.89%) in peak signal-to-noise ratio.

Discussion

The LFT strategy shows great potential to address the issues of "catastrophic forgetting" and overfitting in transfer scenarios for MRI reconstruction, while reducing the reliance on the amount of data in the target domain. Linear fine-tuning is expected to shorten the development cycle of reconstruction models for adapting complicated clinical scenarios, thereby enhancing the clinical applicability of deep MRI reconstruction.

Hierarchical Perception Adversarial Learning Framework for Compressed Sensing MRI

The long acquisition time has limited the accessibility of magnetic resonance imaging (MRI) because it leads to patient discomfort and motion artifacts. Although several MRI techniques have been proposed to reduce the acquisition time, compressed sensing in magnetic resonance imaging (CS-MRI) enables fast acquisition without compromising SNR and resolution. However, existing CS-MRI methods suffer from the challenge of aliasing artifacts. This challenge results in the noise-like textures and missing the fine details, thus leading to unsatisfactory reconstruction performance. To tackle this challenge, we propose a hierarchical perception adversarial learning framework (HP-ALF). HP-ALF can perceive the image information in the hierarchical mechanism: image-level perception and patch-level perception. The former can reduce the visual perception difference in the entire image, and thus achieve aliasing artifact removal. The latter can reduce this difference in the regions of the image, and thus recover fine details. Specifically, HP-ALF achieves the hierarchical mechanism by utilizing multilevel perspective discrimination. This discrimination can provide the information from two perspectives (overall and regional) for adversarial learning. It also utilizes a global and local coherent discriminator to provide structure information to the generator during training. In addition, HP-ALF contains a context-aware learning block to effectively exploit the slice information between individual images for better reconstruction performance. The experiments validated on three datasets demonstrate the effectiveness of HP-ALF and its superiority to the comparative methods.

An artificial bee bare-bone hunger games search for global optimization and high-dimensional feature selection

The domains of contemporary medicine and biology have generated substantial high-dimensional genetic data. Identifying representative genes and decreasing the dimensionality of the data can be challenging. The goal of gene selection is to minimize computing costs and enhance classification precision. Therefore, this article designs a new wrapper gene selection algorithm named artificial bee bare-bone hunger games search (ABHGS), which is the hunger games search (HGS) integrated with an artificial bee strategy and a Gaussian bare-bone structure to address this issue. To evaluate and validate the performance of our proposed method, ABHGS is compared to HGS and a single strategy embedded in HGS, six classic algorithms, and ten advanced algorithms on the CEC 2017 functions. The experimental results demonstrate that the bABHGS outperforms the original HGS. Compared to peers, it increases classification accuracy and decreases the number of selected features, indicating its actual engineering utility in spatial search and feature selection.

An enhanced Runge Kutta boosted machine learning framework for medical diagnosis

With the development and maturity of machine learning methods, medical diagnosis aided with machine learning methods has become a popular method to assist doctors in diagnosing and treating patients. However, machine learning methods are greatly affected by their hyperparameters, for instance, the kernel parameter in kernel extreme learning machine (KELM) and the learning rate in residual neural networks (ResNet). If the hyperparameters are appropriately set, the performance of the classifier can be significantly improved. To boost the performance of the machine learning methods, this paper proposes to improve the Runge Kutta optimizer (RUN) to adaptively adjust the hyperparameters of the machine learning methods for medical diagnosis purposes. Although RUN has a solid mathematical theoretical foundation, there are still some performance defects when dealing with complex optimization problems. To remedy these defects, this paper proposes a new enhanced RUN method with a grey wolf mechanism and an orthogonal learning mechanism called GORUN. The superior performance of the GORUN was validated against other well-established optimizers on IEEE CEC 2017 benchmark functions. Then, the proposed GORUN is employed to optimize the machine learning models, including the KELM and ResNet, to construct robust models for medical diagnosis. The performance of the proposed machine learning framework was validated on several medical data sets, and the experimental results have demonstrated its superiority.

Pneumothorax prediction using a foraging and hunting based ant colony optimizer assisted support vector machine

Although PNLB is generally considered safe, it is still invasive and risky. Pneumothorax, the most common complication of lung puncture, can cause shortness of breath, chest pain, and even life-threatening. Therefore, the auxiliary diagnosis for pneumothorax is of great clinical interest. This paper proposes an ant colony optimizer with slime mould foraging behavior and collaborative hunting, called SCACO, in which slime mould foraging behavior is combined to improve the convergence accuracy and solution quality of ACOR. Then the ability of ACO to jump out of the local optimum is optimized by an adaptive collaborative hunting strategy when trapped in the local optimum. As a first step toward Pneumothorax diagnostic prediction, we suggested an SVM classifier based on bSCACO (bSCACO-SVM), which uses the proposed SCACO's binary version as the basis for its feature selection algorithms. To demonstrate the SCACO performance, we first used the slime mould foraging behavior and adaptive cooperative hunting strategy, then compared SCACO with nine basic algorithms and nine variants, respectively. Finally, we verified bSCACO-SVM on various widely used public datasets and applied it to the Pneumothorax prediction issue, showing that it has robust classification prediction capacity and can be successfully employed for tuberculous pleural effusion diagnostic prediction.

A grade-based search adaptive random slime mould optimizer for lupus nephritis image segmentation

The segmentation of medical images is a crucial and demanding step in medical image processing that offers a solid foundation for subsequent extraction and analysis of medical image data. Although multi-threshold image segmentation is the most used and specialized basic image segmentation technique, it is computationally demanding and often produces subpar segmentation results, hence restricting its application. To solve this issue, this work develops a multi-strategy-driven slime mould algorithm (RWGSMA) for multi-threshold image segmentation. Specifically, the random spare strategy, the double adaptive weigh strategy, and the grade-based search strategy are used to improve the performance of SMA, resulting in an enhanced SMA version. The random spare strategy is mainly used to accelerate the convergence rate of the algorithm. To prevent SMA from falling towards the local optimum, the double adaptive weights are also applied. The grade-based search approach has also been developed to boost convergence performance. This study evaluates the efficacy of RWGSMA from many viewpoints using 30 test suites from IEEE CEC2017 to effectively demonstrate the importance of these techniques in RWGSMA. In addition, numerous typical images were used to show RWGSMA's segmentation performance. Using the multi-threshold segmentation approach with 2D Kapur's entropy as the RWGSMA fitness function, the suggested algorithm was then used to segment instances of lupus nephritis. The experimental findings demonstrate that the suggested RWGSMA beats numerous similar rivals, suggesting that it has a great deal of promise for segmenting histopathological images.

An enhanced ant colony optimizer with Cauchy-Gaussian fusion and novel movement strategy for multi-threshold COVID-19 X-ray image segmentation

The novel coronavirus pneumonia (COVID-19) is a respiratory disease of great concern in terms of its dissemination and severity, for which X-ray imaging-based diagnosis is one of the effective complementary diagnostic methods. It is essential to be able to separate and identify lesions from their pathology images regardless of the computer-aided diagnosis techniques. Therefore, image segmentation in the pre-processing stage of COVID-19 pathology images would be more helpful for effective analysis. In this paper, to achieve highly effective pre-processing of COVID-19 pathological images by using multi-threshold image segmentation (MIS), an enhanced version of ant colony optimization for continuous domains (MGACO) is first proposed. In MGACO, not only a new move strategy is introduced, but also the Cauchy-Gaussian fusion strategy is incorporated. It has been accelerated in terms of convergence speed and has significantly enhanced its ability to jump out of the local optimum. Furthermore, an MIS method (MGACO-MIS) based on MGACO is developed, where it applies the non-local means, 2D histogram as the basis, and employs 2D Kapur’s entropy as the fitness function. To demonstrate the performance of MGACO, we qualitatively analyze it in detail and compare it with other peers on 30 benchmark functions from IEEE CEC2014, which proves that it has a stronger capability of solving problems over the original ant colony optimization for continuous domains. To verify the segmentation effect of MGACO-MIS, we conducted a comparison experiment with eight other similar segmentation methods based on real pathology images of COVID-19 at different threshold levels. The final evaluation and analysis results fully demonstrate that the developed MGACO-MIS is sufficient to obtain high-quality segmentation results in the COVID-19 image segmentation and has stronger adaptability to different threshold levels than other methods. Therefore, it has been well-proven that MGACO is an excellent swarm intelligence optimization algorithm, and MGACO-MIS is also an excellent segmentation method.

HIWDNet: A hybrid image-wavelet domain network for fast magnetic resonance image reconstruction

The application of Magnetic Resonance Imaging (MRI) is limited due to the long acquisition time of k-space signals. Recently, many deep learning-based MR image reconstruction methods have been proposed to reduce acquisition time and improve MRI image quality by reconstructing images from under-sampled k-space data. However, these methods suffer from two shortcomings. Firstly, the reconstruction network are mainly designed in the image domain or frequency domain, while ignoring the characteristics of time-frequency features in the wavelet domain. In addition, the existing cross-domain methods design the same reconstruction network in different transform domains, so that the network cannot learn targeted information for different domains. To solve the above problems, we propose a Hybrid Image-Wavelet Domain Reconstruction Network (HIWDNet) for fast MRI reconstruction. Specifically, we employ Cross-scale Dense Feature Fusion Module (CDFFM) in the image domain to reconstruct the basic structure of MR images, while introducing Region Adaptive Artifact Removal Module (RAARM) to remove aliasing artifacts in large areas. Then, a Wavelet Sub-band Reconstruction Module (WSRM) is proposed to refine wavelet sub-bands to improve the accuracy of HIWDNet. The proposed method is evaluated in different sampling modes on the fastMRI dataset, the CC359 dataset and the IXI dataset. Extensive experimental results show that HIWDNet achieves better results on both SSIM and PSNR evaluation metrics compared with other methods.

A Systematic Literature Review on Applications of GAN-Synthesized Images for Brain MRI

With the advances in brain imaging, magnetic resonance imaging (MRI) is evolving as a popular radiological tool in clinical diagnosis. Deep learning (DL) methods can detect abnormalities in brain images without an extensive manual feature extraction process. Generative adversarial network (GAN)-synthesized images have many applications in this field besides augmentation, such as image translation, registration, super-resolution, denoising, motion correction, segmentation, reconstruction, and contrast enhancement. The existing literature was reviewed systematically to understand the role of GAN-synthesized dummy images in brain disease diagnosis. Web of Science and Scopus databases were extensively searched to find relevant studies from the last 6 years to write this systematic literature review (SLR). Predefined inclusion and exclusion criteria helped in filtering the search results. Data extraction is based on related research questions (RQ). This SLR identifies various loss functions used in the above applications and software to process brain MRIs. A comparative study of existing evaluation metrics for GAN-synthesized images helps choose the proper metric for an application. GAN-synthesized images will have a crucial role in the clinical sector in the coming years, and this paper gives a baseline for other researchers in the field.

Fast MRI Reconstruction: How Powerful Transformers Are?

Magnetic resonance imaging (MRI) is a widely used non-radiative and non-invasive method for clinical interro-gation of organ structures and metabolism, with an inherently long scanning time. Methods by k-space undersampling and deep learning based reconstruction have been popularised to accelerate the scanning process. This work focuses on investigating how powerful transformers are for fast MRI by exploiting and comparing different novel network architectures. In particular, a generative adversarial network (GAN) based Swin transformer (ST-GAN) was introduced for the fast MRI reconstruction. To further preserve the edge and texture information, edge enhanced GAN based Swin transformer (EES-GAN) and texture enhanced GAN based Swin transformer (TES-GAN) were also developed, where a dual-discriminator GAN structure was applied. We compared our proposed GAN based transformers, standalone Swin transformer and other convolutional neural networks based GAN model in terms of the evaluation metrics PSNR, SSIM and FID. We showed that transformers work well for the MRI reconstruction from different undersampling conditions. The utilisation of GAN's adversarial structure improves the quality of images reconstructed when undersampled for 30% or higher. The code is publicly available at https://github.comJayanglab/SwinGANMR.

An Enhanced Priori Knowledge GAN for CT Images Generation of Early Lung Nodules with Small-Size Labelled Samples

The small size of labelled samples is one of the challenging problems in identifying early lung nodules from CT images using deep learning methods. Recent literature on the topic shows that deep convolutional generative adversarial network (DCGAN) has been used in medical data synthesis and gained some success, but does not demonstrate satisfactory results in synthesizing CT images. It primarily suffers from the problem of model convergence and is prone to mode collapse. In this paper, we propose a generative adversarial network (GAN) model with prior knowledge to generate CT images of early lung nodules from a small-size of labelled samples, i.e., SLS-PriGAN. Particularly, a knowledge acquisition network and a sharpening network are designed for priori knowledge learning and acquisition, and then, a GAN model is developed to produce CT images of early lung nodules. To validate our method, a general fast R-CNN network is trained using the CT images generated by SLS-PriGAN. The experiment result shows that it achieved a recognizing accuracy of 91%, a recall rate of 81%, and F1 score of 0.85 in identifying clinic CT images of early lung nodules. This provides a promising way of identifying early lung nodules from CT images using deep learning with small-size labelled samples.

The role of generative adversarial networks in brain MRI: a scoping review

The performance of artificial intelligence (AI) for brain MRI can improve if enough data are made available. Generative adversarial networks (GANs) showed a lot of potential to generate synthetic MRI data that can capture the distribution of real MRI. Besides, GANs are also popular for segmentation, noise removal, and super-resolution of brain MRI images. This scoping review aims to explore how GANs methods are being used on brain MRI data, as reported in the literature. The review describes the different applications of GANs for brain MRI, presents the most commonly used GANs architectures, and summarizes the publicly available brain MRI datasets for advancing the research and development of GANs-based approaches. This review followed the guidelines of PRISMA-ScR to perform the study search and selection. The search was conducted on five popular scientific databases. The screening and selection of studies were performed by two independent reviewers, followed by validation by a third reviewer. Finally, the data were synthesized using a narrative approach. This review included 139 studies out of 789 search results. The most common use case of GANs was the synthesis of brain MRI images for data augmentation. GANs were also used to segment brain tumors and translate healthy images to diseased images or CT to MRI and vice versa. The included studies showed that GANs could enhance the performance of AI methods used on brain MRI imaging data. However, more efforts are needed to transform the GANs-based methods in clinical applications.

Transfer Learning Improving Predictive Mortality Models for Patients in End-Stage Renal Disease

Deep learning is becoming a fundamental piece in the paradigm shift from evidence-based to data-based medicine. However, its learning capacity is rarely exploited when working with small data sets. Through transfer learning (TL), information from a source domain is transferred to a target one to enhance a learning task in such domain. The proposed TL mechanisms are based on sample and feature space augmentation. Thus, deep autoencoders extract complex representations for the data in the TL approach. Their latent representations, the so-called codes, are handled to transfer information among domains. The transfer of samples is carried out by computing a latent space mapping matrix that links codes from both domains for later reconstruction. The feature space augmentation is based on the computation of the average of the most similar codes from one domain. Such an average augments the features in a target domain. The proposed framework is evaluated in the prediction of mortality in patients in end-stage renal disease, transferring information related to the mortality of patients with acute kidney injury from the massive database MIMIC-III. Compared to other TL mechanisms, the proposed approach improves 6–11% in previous mortality predictive models. The integration of TL approaches into learning tasks in pathologies with data volume issues could encourage the use of data-based medicine in a clinical setting.

Generative adversarial network-based post-processed image super-resolution technology for accelerating brain MRI: comparison with compressed sensing

BACKGROUND

It is unclear whether deep-learning-based super-resolution technology (SR) or compressed sensing technology (CS) can accelerate magnetic resonance imaging (MRI) .

PURPOSE

To compare SR accelerated images with CS images regarding the image similarity to reference 2D- and 3D gradient-echo sequence (GRE) brain MRI.

MATERIAL AND METHODS

We prospectively acquired 1.3× and 2.0× faster 2D and 3D GRE images of 20 volunteers from the reference time by reducing the matrix size or increasing the CS factor. For SR, we trained the generative adversarial network (GAN), upscaling the low-resolution images to the reference images with twofold cross-validation. We compared the structural similarity (SSIM) index of accelerated images to the reference image. The rate of incorrect answers of a radiologist discriminating faster and reference image was used as a subjective image similarity (ISM) index.

RESULTS

The SR demonstrated significantly higher SSIM than the CS (SSIM=0.9993-0.999 vs. 0.9947-0.9986; P < 0.001). In 2D GRE, it was challenging to discriminate the SR image from the reference image, compared to the CS (ISM index 40% vs. 17.5% in 1.3×; P = 0.039 and 17.5% vs. 2.5% in 2.0×; P = 0.034). In 3D GRE, the CS revealed a significantly higher ISM index than the SR (22.5% vs. 2.5%; P = 0.011) in 2.0 × faster images. However, the ISM index was identical for the 2.0× CS and 1.3× SR (22.5% vs. 27.5%; P = 0.62) with comparable time costs.

CONCLUSION

The GAN-based SR outperformed CS in image similarity with 2D GRE for MRI acceleration. In addition, CS was more advantageous in 3D GRE than SR.

Edge-enhanced dual discriminator generative adversarial network for fast MRI with parallel imaging using multi-view information

In clinical medicine, magnetic resonance imaging (MRI) is one of the most important tools for diagnosis, triage, prognosis, and treatment planning. However, MRI suffers from an inherent slow data acquisition process because data is collected sequentially in k-space. In recent years, most MRI reconstruction methods proposed in the literature focus on holistic image reconstruction rather than enhancing the edge information. This work steps aside this general trend by elaborating on the enhancement of edge information. Specifically, we introduce a novel parallel imaging coupled dual discriminator generative adversarial network (PIDD-GAN) for fast multi-channel MRI reconstruction by incorporating multi-view information. The dual discriminator design aims to improve the edge information in MRI reconstruction. One discriminator is used for holistic image reconstruction, whereas the other one is responsible for enhancing edge information. An improved U-Net with local and global residual learning is proposed for the generator. Frequency channel attention blocks (FCA Blocks) are embedded in the generator for incorporating attention mechanisms. Content loss is introduced to train the generator for better reconstruction quality. We performed comprehensive experiments on Calgary-Campinas public brain MR dataset and compared our method with state-of-the-art MRI reconstruction methods. Ablation studies of residual learning were conducted on the MICCAI13 dataset to validate the proposed modules. Results show that our PIDD-GAN provides high-quality reconstructed MR images, with well-preserved edge information. The time of single-image reconstruction is below 5ms, which meets the demand of faster processing.

CryoETGAN: Cryo-Electron Tomography Image Synthesis via Unpaired Image Translation

Cryo-electron tomography (Cryo-ET) has been regarded as a revolution in structural biology and can reveal molecular sociology. Its unprecedented quality enables it to visualize cellular organelles and macromolecular complexes at nanometer resolution with native conformations. Motivated by developments in nanotechnology and machine learning, establishing machine learning approaches such as classification, detection and averaging for Cryo-ET image analysis has inspired broad interest. Yet, deep learning-based methods for biomedical imaging typically require large labeled datasets for good results, which can be a great challenge due to the expense of obtaining and labeling training data. To deal with this problem, we propose a generative model to simulate Cryo-ET images efficiently and reliably: CryoETGAN. This cycle-consistent and Wasserstein generative adversarial network (GAN) is able to generate images with an appearance similar to the original experimental data. Quantitative and visual grading results on generated images are provided to show that the results of our proposed method achieve better performance compared to the previous state-of-the-art simulation methods. Moreover, CryoETGAN is stable to train and capable of generating plausibly diverse image samples.

MAGnitude-Image-to-Complex K-space (MAGIC-K) Net: A Data Augmentation Network for Image Reconstruction

Deep learning has demonstrated superior performance in image reconstruction compared to most conventional iterative algorithms. However, their effectiveness and generalization capability are highly dependent on the sample size and diversity of the training data. Deep learning-based reconstruction requires multi-coil raw k-space data, which are not collected by routine scans. On the other hand, large amounts of magnitude images are readily available in hospitals. Hence, we proposed the MAGnitude Images to Complex K-space (MAGIC-K) Net to generate multi-coil k-space data from existing magnitude images and a limited number of required raw k-space data to facilitate the reconstruction. Compared to some basic data augmentation methods applying global intensity and displacement transformations to the source images, the MAGIC-K Net can generate more realistic intensity variations and displacements from pairs of anatomical Digital Imaging and Communications in Medicine (DICOM) images. The reconstruction performance was validated in 30 healthy volunteers and 6 patients with different types of tumors. The experimental results demonstrated that the high-resolution Diffusion Weighted Image (DWI) reconstruction benefited from the proposed augmentation method. The MAGIC-K Net enabled the deep learning network to reconstruct images with superior performance in both healthy and tumor patients, qualitatively and quantitatively.