Labelling with dynamics: A data-efficient learning paradigm for medical image segmentation.

The success of deep learning on image classification and recognition tasks has led to new applications in diverse contexts, including the field of medical imaging. However, two properties of deep neural networks (DNNs) may limit their future use in medical applications. The first is that DNNs require a large amount of labeled training data, and the second is that the deep learning-based models lack interpretability. In this paper, we propose and investigate a data-efficient framework for the task of general medical image segmentation. We address the two aforementioned challenges by introducing domain knowledge in the form of a strong prior into a deep learning framework. This prior is expressed by a customized dynamical system. We performed experiments on two different datasets, namely JSRT and ISIC2016 (heart and lungs segmentation on chest X-ray images and skin lesion segmentation on dermoscopy images). We have achieved competitive results using the same amount of training data compared to the state-of-the-art methods. More importantly, we demonstrate that our framework is extremely data-efficient, and it can achieve reliable results using extremely limited training data. Furthermore, the proposed method is rotationally invariant and insensitive to initialization.

Description and reconstruction of typical structured light beams with vector spherical wave functions.

It is well known that the generalized Lorenz-Mie theory (GLMT) is a rigorous analytical method for dealing with the interaction between light beams and spherical particles, which involves the description and reconstruction of the light beams with vector spherical wave functions (VSWFs). In this paper, a detailed study on the description and reconstruction of the typical structured light beams with VSWFs is reported. We first systematically derive the so-called beam shape coefficients (BSCs) of typical structured light beams, including the fundamental Gaussian beam, Hermite-Gaussian beam, Laguerre-Gaussian beam, Bessel beam, and Airy beam, with the aid of the angular spectrum decomposition method. Then based on the derived BSCs, we reconstruct these structured light beams using VSWFs and compare the results of the reconstructed beams with those of the original beams. Our results will be useful in the study of the interaction of typical structured light beams with spherical particles in the framework of GLMT.

Multi-Source Domain Adaptation for Medical Image Segmentation.

Unsupervised domain adaptation(UDA) aims to mitigate the performance drop of models tested on the target domain, due to the domain shift from the target to sources. Most UDA segmentation methods focus on the scenario of solely single source domain. However, in practical situations data with gold standard could be available from multiple sources (domains), and the multi-source training data could provide more information for knowledge transfer. How to utilize them to achieve better domain adaptation yet remains to be further explored. This work investigates multi-source UDA and proposes a new framework for medical image segmentation. Firstly, we employ a multi-level adversarial learning scheme to adapt features at different levels between each of the source domains and the target, to improve the segmentation performance. Then, we propose a multi-model consistency loss to transfer the learned multi-source knowledge to the target domain simultaneously. Finally, we validated the proposed framework on two applications, i.e., multi-modality cardiac segmentation and cross-modality liver segmentation. The results showed our method delivered promising performance and compared favorably to state-of-the-art approaches.

Effectiveness of Huangqi Guizhi Wuwu decoction combined with rehabilitation training for shoulder hand syndrome after stroke: A systematic review and meta-analysis.

BACKGROUND: Shoulder hand syndrome (SHS) is a common complication of stroke. This meta-analysis aimed to evaluate the effectiveness of Huangqi Guizhi Wuwu decoction (HGWD) combined with rehabilitation training in managing it, as its efficacy remains inconclusive. METHODS: Seven databases, including PubMed, EMBASE, Cochrane Library, SinoMed, Chinese National Knowledge Infrastructure, Wanfang Data, and VIP database were searched in this study. The search deadline was April 30, 2023. Randomized controlled trials that included either standalone rehabilitation training or HGWD combined with rehabilitation training were included, and data were independently extracted by 2 reviewers who assessed the risk of bias. RESULTS: Thirteen studies involving 1270 patients were included in this study. Meta-analysis showed that the combined treatment was significantly more effective than standalone rehabilitation therapy (odds ratio = 4.49; 95%CI: 2.98-6.76; Z = 7.17; P < .00001). Compared with the control group, the intervention group had a lower visual analog scale score (mean difference [MD] = -2.80, 95%CI (-3.15, -2.45), Z = 15.84, P < .00001). In addition, the Fugl-Meyer assessment scale score improved (MD = 9.69, 95%CI (7.60, 11.78), Z = 9.08, P < .00001). The SHS score in the intervention group decreased more compared to the control group (standard mean difference = -2.27, 95%CI (-3.19, -1.34), Z = 4.79, P < .00001). Serum biomarkers related to SHS decreased, including serum substance P (MD = -7.52, 95%CI (-8.55, -6.48), Z = 14, P < .00001) and bradykinin (MD = -1.81, 95%CI (-2.68, -0.95), Z = 4.1, P < .00001). Although there was no statistical difference in joint mobility score (MD = -4.19, 95%CI (-8.16, -0.22), Z = 4.79, P = .28), sensitivity analysis after excluding one study still suggested that the joint mobility score of the combined treatment group was higher than that of the standalone rehabilitation treatment group. CONCLUSION: The results of this study indicate that HGWD combined with rehabilitation training may be more effective in treating SHS after stroke compared to standalone rehabilitation therapy.

Multi-target landmark detection with incomplete images via reinforcement learning and shape prior embedding.

Medical images are generally acquired with limited field-of-view (FOV), which could lead to incomplete regions of interest (ROI), and thus impose a great challenge on medical image analysis. This is particularly evident for the learning-based multi-target landmark detection, where algorithms could be misleading to learn primarily the variation of background due to the varying FOV, failing the detection of targets. Based on learning a navigation policy, instead of predicting targets directly, reinforcement learning (RL)-based methods have the potential to tackle this challenge in an efficient manner. Inspired by this, in this work we propose a multi-agent RL framework for simultaneous multi-target landmark detection. This framework is aimed to learn from incomplete or (and) complete images to form an implicit knowledge of global structure, which is consolidated during the training stage for the detection of targets from either complete or incomplete test images. To further explicitly exploit the global structural information from incomplete images, we propose to embed a shape model into the RL process. With this prior knowledge, the proposed RL model can not only localize dozens of targets simultaneously, but also work effectively and robustly in the presence of incomplete images. We validated the applicability and efficacy of the proposed method on various multi-target detection tasks with incomplete images from practical clinics, using body dual-energy X-ray absorptiometry (DXA), cardiac MRI and head CT datasets. Results showed that our method could predict whole set of landmarks with incomplete training images up to 80% missing proportion (average distance error 2.29 cm on body DXA), and could detect unseen landmarks in regions with missing image information outside FOV of target images (average distance error 6.84 mm on 3D half-head CT). Our code will be released via https://zmiclab.github.io/projects.html.

MyoPS: A benchmark of myocardial pathology segmentation combining three-sequence cardiac magnetic resonance images.

Assessment of myocardial viability is essential in diagnosis and treatment management of patients suffering from myocardial infarction, and classification of pathology on the myocardium is the key to this assessment. This work defines a new task of medical image analysis, i.e., to perform myocardial pathology segmentation (MyoPS) combining three-sequence cardiac magnetic resonance (CMR) images, which was first proposed in the MyoPS challenge, in conjunction with MICCAI 2020. Note that MyoPS refers to both myocardial pathology segmentation and the challenge in this paper. The challenge provided 45 paired and pre-aligned CMR images, allowing algorithms to combine the complementary information from the three CMR sequences for pathology segmentation. In this article, we provide details of the challenge, survey the works from fifteen participants and interpret their methods according to five aspects, i.e., preprocessing, data augmentation, learning strategy, model architecture and post-processing. In addition, we analyze the results with respect to different factors, in order to examine the key obstacles and explore the potential of solutions, as well as to provide a benchmark for future research. The average Dice scores of submitted algorithms were 0.614±0.231 and 0.644±0.153 for myocardial scars and edema, respectively. We conclude that while promising results have been reported, the research is still in the early stage, and more in-depth exploration is needed before a successful application to the clinics. MyoPS data and evaluation tool continue to be publicly available upon registration via its homepage (www.sdspeople.fudan.edu.cn/zhuangxiahai/0/myops20/).

Water quality characteristics and evaluation of Qilian Mountain National Park section in Heihe River Basin based on water quality indices and 3D fluorescence technology.

The water quality of the Heihe River Basin affects the life quality and health of tens of thousands of residents along it. However, there are relatively few studies that evaluate its water quality. In this study, we used principal component analysis (PCA), an improved comprehensive water quality index (WQI), and three-dimensional (3D) fluorescence technology to identify pollutants and evaluate water quality at nine monitoring sites in the Qilian Mountain National Park in Heihe River Basin. PCA was applied to concentrate the water quality indices into nine items. The analysis shows that the water quality in the study area is mainly polluted by organic matter, nitrogen, and phosphorus. According to the revised WQI model, the water quality of the study area is from moderate to good, while the water quality of Qinghai section is worse than that of Gansu section. According to the 3D fluorescence spectrum analysis of the monitoring sites, the organic pollution of water comes from vegetation decay, animal feces, and some human activities. This study can not only provide support and basis for water environment protection and management in the Heihe River Basin, but also promote the healthy development of the water environment in the Qilian Mountains.

Minimizing Estimated Risks on Unlabeled Data: A New Formulation for Semi-Supervised Medical Image Segmentation.

Supervised segmentation can be costly, particularly in applications of biomedical image analysis where large scale manual annotations from experts are generally too expensive to be available. Semi-supervised segmentation, able to learn from both the labeled and unlabeled images, could be an efficient and effective alternative for such scenarios. In this work, we propose a new formulation based on risk minimization, which makes full use of the unlabeled images. Different from most of the existing approaches which solely explicitly guarantee the minimization of prediction risks from the labeled training images, the new formulation also considers the risks on unlabeled images. Particularly, this is achieved via an unbiased estimator, based on which we develop a general framework for semi-supervised image segmentation. We validate this framework on three medical image segmentation tasks, namely cardiac segmentation on ACDC2017, optic cup and disc segmentation on REFUGE dataset and 3D whole heart segmentation on MM-WHS dataset. Results show that the proposed estimator is effective, and the segmentation method achieves superior performance and demonstrates great potential compared to the other state-of-the-art approaches. Our code and data will be released via https://zmiclab.github.io/projects.html, once the manuscript is accepted for publication.

A New Framework of Swarm Learning Consolidating Knowledge From Multi-Center Non-IID Data for Medical Image Segmentation.

Large training datasets are important for deep learning-based methods. For medical image segmentation, it could be however difficult to obtain large number of labeled training images solely from one center. Distributed learning, such as swarm learning, has the potential to use multi-center data without breaching data privacy. However, data distributions across centers can vary a lot due to the diverse imaging protocols and vendors (known as feature skew). Also, the regions of interest to be segmented could be different, leading to inhomogeneous label distributions (referred to as label skew). With such non-independently and identically distributed (Non-IID) data, the distributed learning could result in degraded models. In this work, we propose a novel swarm learning approach, which assembles local knowledge from each center while at the same time overcomes forgetting of global knowledge during local training. Specifically, the approach first leverages a label skew-awared loss to preserve the global label knowledge, and then aligns local feature distributions to consolidate global knowledge against local feature skew. We validated our method in three Non-IID scenarios using four public datasets, including the Multi-Centre, Multi-Vendor and Multi-Disease Cardiac Segmentation (M&Ms) dataset, the Federated Tumor Segmentation (FeTS) dataset, the Multi-Modality Whole Heart Segmentation (MMWHS) dataset and the Multi-Site Prostate T2-weighted MRI segmentation (MSProsMRI) dataset. Results show that our method could achieve superior performance over existing methods. Code will be released via https://zmiclab.github.io/projects.html once the paper gets accepted.

Chirality of optical vortex beams reflected from an air-chiral medium interface.

Chirality plays an important role in understanding of the chiral light-matter interaction. In this work, we study theoretically and numerically the chirality of optical vortex beams reflected from an air-chiral medium interface. A theoretical model that takes into full account the vectorial nature of electromagnetic fields is developed to describe the reflection of optical vortex beams at an interface between air and a chiral medium. Some numerical simulations are performed and discussed. The results show that the chirality of the reflected vortex beams can be well controlled by the relative chiral parameter of the medium and is significantly affected by the incidence angle, topological charge, and polarization state of the incident beam. Our results provide new, to the best of our knowledge, insights into the interactions between optical vortex beams with chiral matter, and may have potential application in optical chirality manipulation.

Scattering analysis of two-dimensional Airy beams by typical non-spherical particles.

The scattering of structured light beams by various particles is an important subject of research with myriad practical applications, such as the manipulation, measurement, and diagnosis of small particles. We carry out an analysis of the scattering of two-dimensional (2D) Airy beams by typical non-spherical particles. The electric and magnetic field vectors of the incident Airy beams are derived by introducing a vector potential in the Lorenz gauge. The scattered fields of the particles are obtained by utilizing the method of moments based on surface integral equations. Some numerical simulations for the scattering of 2D Airy beams by several selected non-spherical particles are performed and analyzed. Especially, a spheroidal particle is taken as an example, and the effects of various parameters describing the 2D Airy beams on its differential scattering cross section are examined. It is expected that this work will be helpful for understanding the interactions of 2D Airy beams with non-spherical particles and their further applications.

Echogenic swirling pattern, carcinoembryonic antigen, and lactate dehydrogenase in the diagnosis of malignant pleural effusion.

The echogenic swirling pattern has a role in predicting malignant pleural effusion (MPE). However, its predictive ability is suboptimal, and its clinical utility remains to be defined. The aim of this study was to assess the diagnostic potential of the echogenic swirling pattern combined with pleural carcinoembryonic antigen (CEA) and routine laboratory tests of pleural effusion in MPE. The 80 consecutive patients with underlying malignancy and pleural effusions were recruited. All patients underwent one diagnostic thoracentesis with a cytologic examination of pleural fluid. Our study showed that the sensitivity of echogenic swirling patterns in MPE diagnosis was 67.7%, specificity was 72.2%, positive predictive value (PPV) was 89.4%, and negative predictive value (NPV) was 39.4%. Both CEA and lactate dehydrogenase (LDH) had acceptable sensitivity (71.0% and 60.7%) and specificity (72.2% and 77.8%). Combining the echogenic swirling pattern, pleural CEA, and pleural LDH, the highest sensitivity (95.2%) with a good PPV (86.8) was reached. In this clinical study, we found that combining the echogenic swirling pattern, pleural CEA, and pleural LDH had a higher sensitivity and a high positive predictive value for the diagnosis of MPE. This combination is a potentially suitable method for MPE screening in cancer patients with pleural effusions.

Hydrothermal synthesis of magnetic nano-CoFe2O4 catalyst and its enhanced degradation of amoxicillin by activated permonosulfate.

Advanced oxidation process (AOP) has attracted widespread attention because it can effectively remove antibiotics in water, but its practical engineering application is limited by the problems of the low efficiency and difficult recovery of the catalyst. In the study, nano-spinel CoFe2O4 was prepared by hydrothermal method and served as the peroxymonosulfate (PMS) catalyst to degrade antibiotic amoxicillin (AMX). The reaction parameters such as CoFe2O4 dosage, AMX concentration, and initial pH value were also optimized. The reaction mechanism was proposed through free radical capture experiment and possible degradation pathway analysis. In addition, the magnetic recovery performance and stability of the catalyst were evaluated. Results showed that 85.5% of AMX could be removed within 90 min at optimal conditions. Sulfate radicals and hydroxyl radicals were the active species for AMX degradation. Moreover, the catalyst showed excellent magnetism and stability in the cycle experiment, which has great potential in the AOP treatment of antibiotic polluted wastewater.

Light scattering of Laguerre-Gaussian vortex beams by arbitrarily shaped chiral particles.

Laguerre-Gaussian (LG) beams with vortex phase possess a handedness, which would produce chiroptical interactions with chiral matter and may be used to probe structural chirality of matter. In this paper, we numerically investigate the light scattering of LG vortex beams by chiral particles. Using the vector potential method, the electric and magnetic field components of the incident LG vortex beams are derived. The method of moments (MoM) based on surface integral equations (SIEs) is applied to solve the scattering problems involving arbitrarily shaped chiral particles. The numerical results for the differential scattering cross sections (DSCSs) of several selected chiral particles illuminated by LG vortex beams are presented and analyzed. In particular, we show how the DSCSs depend on the chiral parameter of the particles and on the parameters describing the incident LG vortex beams, including the topological charge, the state of circular polarization, and the beam waist. This research may provide useful insights into the interaction of vortex beams with chiral particles and its further applications.

Unsupervised Domain Adaptation With Variational Approximation for Cardiac Segmentation.

Unsupervised domain adaptation is useful in medical image segmentation. Particularly, when ground truths of the target images are not available, domain adaptation can train a target-specific model by utilizing the existing labeled images from other modalities. Most of the reported works mapped images of both the source and target domains into a common latent feature space, and then reduced their discrepancy either implicitly with adversarial training or explicitly by directly minimizing a discrepancy metric. In this work, we propose a new framework, where the latent features of both domains are driven towards a common and parameterized variational form, whose conditional distribution given the image is Gaussian. This is achieved by two networks based on variational auto-encoders (VAEs) and a regularization for this variational approximation. Both of the VAEs, each for one domain, contain a segmentation module, where the source segmentation is trained in a supervised manner, while the target one is trained unsupervisedly. We validated the proposed domain adaptation method using two cardiac segmentation tasks, i.e., the cross-modality (CT and MR) whole heart segmentation and the cross-sequence cardiac MR segmentation. Results show that the proposed method achieved better accuracies compared to two state-of-the-art approaches and demonstrated good potential for cardiac segmentation. Furthermore, the proposed explicit regularization was shown to be effective and efficient in narrowing down the distribution gap between domains, which is useful for unsupervised domain adaptation. The code and data have been released via https://zmiclab.github.io/projects.html.

Disentangle domain features for cross-modality cardiac image segmentation.

Unsupervised domain adaptation (UDA) generally learns a mapping to align the distribution of the source domain and target domain. The learned mapping can boost the performance of the model on the target data, of which the labels are unavailable for model training. Previous UDA methods mainly focus on domain-invariant features (DIFs) without considering the domain-specific features (DSFs), which could be used as complementary information to constrain the model. In this work, we propose a new UDA framework for cross-modality image segmentation. The framework first disentangles each domain into the DIFs and DSFs. To enhance the representation of DIFs, self-attention modules are used in the encoder which allows attention-driven, long-range dependency modeling for image generation tasks. Furthermore, a zero loss is minimized to enforce the information of target (source) DSFs, contained in the source (target) images, to be as close to zero as possible. These features are then iteratively decoded and encoded twice to maintain the consistency of the anatomical structure. To improve the quality of the generated images and segmentation results, several discriminators are introduced for adversarial learning. Finally, with the source data and their DIFs, we train a segmentation network, which can be applicable to target images. We validated the proposed framework for cross-modality cardiac segmentation using two public datasets, and the results showed our method delivered promising performance and compared favorably to state-of-the-art approaches in terms of segmentation accuracies. The source code of this work will be released via https://zmiclab.github.io/projects.html, once this manuscript is accepted for publication.

CF Distance: A New Domain Discrepancy Metric and Application to Explicit Domain Adaptation for Cross-Modality Cardiac Image Segmentation.

Domain adaptation has great values in unpaired cross-modality image segmentation, where the training images with gold standard segmentation are not available from the target image domain. The aim is to reduce the distribution discrepancy between the source and target domains. Hence, an effective measurement for this discrepancy is critical. In this work, we propose a new metric based on characteristic functions of distributions. This metric, referred to as CF distance, enables explicit domain adaptation, in contrast to the implicit manners minimizing domain discrepancy via adversarial training. Based on this CF distance, we propose an unsupervised domain adaptation framework for cross-modality cardiac segmentation, which consists of image reconstruction and prior distribution matching. We validated the method on two tasks, i.e., the CT-MR cross-modality segmentation and the multi-sequence cardiac MR segmentation. Results showed that the proposed explicit metric was effective in domain adaptation, and the segmentation method delivered promising and superior performance, compared to other state-of-the-art techniques. The data and source code of this work has been released via https://zmiclab.github.io/projects.html.

Waste Brick as Constructed Wetland Fillers to Treat the Tail Water of Sewage Treatment Plant.

Adopting the concept of "using waste to treat waste", the waste bricks will be used for constructed wetland filling. Integrated vertical-flow constructed wetland (IVCW) studied on the purification effect in influent water under three hydraulic loads (0.15, 0.25, 0.35 m/day). The results show that the waste bricks can be used as the carrier for the growth of the system biofilm, and have positive effects on the removal of pollutants in the influent water. Under three different hydraulic load conditions, the vertical flow of CWs can significantly reduce the load of water intake. In the low hydraulic load condition of 0.15 m/day, the average removal rates of chemical oxygen demand (COD), ammonia nitrogen (NH3-N), total nitrogen (TN), and total phosphorus (TP) can reach 66.52%, 72.10%, 56.53% and 91.55% in this system, respectively. The influent pool on removal efficiency of pollutants was obviously higher than that of the upper pool, especially in the inlet surface 0-30 cm ranges. This research has achieved the effect of using "waste" to treat wastewater, which has strong practical significance and popularization value.

Atrial scar quantification via multi-scale CNN in the graph-cuts framework.

Late gadolinium enhancement magnetic resonance imaging (LGE MRI) appears to be a promising alternative for scar assessment in patients with atrial fibrillation (AF). Automating the quantification and analysis of atrial scars can be challenging due to the low image quality. In this work, we propose a fully automated method based on the graph-cuts framework, where the potentials of the graph are learned on a surface mesh of the left atrium (LA) using a multi-scale convolutional neural network (MS-CNN). For validation, we have included fifty-eight images with manual delineations. MS-CNN, which can efficiently incorporate both the local and global texture information of the images, has been shown to evidently improve the segmentation accuracy of the proposed graph-cuts based method. The segmentation could be further improved when the contribution between the t-link and n-link weights of the graph is balanced. The proposed method achieves a mean accuracy of 0.856 ± 0.033 and mean Dice score of 0.702 ± 0.071 for LA scar quantification. Compared to the conventional methods, which are based on the manual delineation of LA for initialization, our method is fully automatic and has demonstrated significantly better Dice score and accuracy (p < 0.01). The method is promising and can be potentially useful in diagnosis and prognosis of AF.

[Microbial Community Structure and Diversity in Cellar Water by 16S rRNA High-throughput Sequencing].

This study was aimed to explore the bacterial diversity of cellar water as well as to study the relationship between the bacterial diversity and environmental factors. The MiSeq high-throughput sequencing was used to analyze and compare the bacterial diversity and community composition of samples from different cellar water samples. Overall 1605 optimized reads were obtained from four samples based on high-throughput sequencing of the V4 region of the 16S rRNA gene. Bacterial species detected in these samples covered 22 phyla,42 classes,71 orders,115 families, 146 genera. Analysis showed that the bacterial diversity was very high in these samples, and there were differences among different samples. The distribution characteristics of the dominant bacteria showed patterns of a large number of rare species and a few common types. Taxonomic assignment analysis indicated that Bacteroidetes,Proteobacteria,Actinobacteria,Verrucomicrobia,OD1 dominated in the Cellar water, and accounted for 87.1% to 94.8% at phylum level. The predominant groups were Actinobacteria,Acidimicrobiia,Cytophagia, Flavobacteriia, Sphingobacteriia,α-Proteobacteria,ß-Proteobacteria,γ-Proteobacteria,Opitutae, Verrucomicrobiae,Pedosphaerae and ZB2 at class level. At genus level Rhodobacter,Dechloromonas,Flavobacterium,Acinetobacter,Comamonas,Pseudomonas,Hydrogenophaga,et al were the abundant taxa, which were mainly denitrifying bacteria and heterotrophic nitrification-aerobic denitrification bacteria. The result of RDA suggested that the influences of different environmental factors on different microbes were different. Bacterial community Ⅱ had significant positive correlation with UV254,permanganate index,BOD5,and Bacterial community Ⅲ had significant positive correlation with TN,NO2--N,NO3--N,TP,NH4+-N. This research should deepen the understanding on microbial community in Cellar water, and provide references for the association of bacterial composition and diversity with environmental factors.