Self-Supervised Node Classification with Strategy and Actively Selected Labeled Set.

To alleviate the impact of insufficient labels in less-labeled classification problems, self-supervised learning improves the performance of graph neural networks (GNNs) by focusing on the information of unlabeled nodes. However, none of the existing self-supervised pretext tasks perform optimally on different datasets, and the choice of hyperparameters is also included when combining self-supervised and supervised tasks. To select the best-performing self-supervised pretext task for each dataset and optimize the hyperparameters with no expert experience needed, we propose a novel auto graph self-supervised learning framework and enhance this framework with a one-shot active learning method. Experimental results on three real world citation datasets show that training GNNs with automatically optimized pretext tasks can achieve or even surpass the classification accuracy obtained with manually designed pretext tasks. On this basis, compared with using randomly selected labeled nodes, using actively selected labeled nodes can further improve the classification performance of GNNs. Both the active selection and the automatic optimization contribute to semi-supervised node classification.

Transcriptomics and flow cytometry reveals the cytotoxicity of aflatoxin B1 and aflatoxin M1 in bovine mammary epithelial cells.

Aflatoxin is a known mycotoxin that pollutes various grains widely in the environment. Aflatoxin B1 (AFB1) and Aflatoxin M1 (AFM1) have been shown to induce cytotoxicity in many cells, yet their effects on mammary epithelial cells remain unclear. In this study, we examined the toxicity and the effects of AFB1 and AFM1 on bovine mammary epithelial cells (BME cells). The cells were treated with AFB1 or AFM1 at a concentration of 0-10 mg/L for 24 or 48 h, followed by cytotoxicity assays, flow cytometry, and transcriptomics. Our results demonstrated that AFB1 and AFM1 induced cell proliferation inhibition, apoptosis and cell cycle arrest. However, the level of intracellular reactive oxygen species has no significant difference. The RNA-Seq results also showed that AFB1 and AFM1 changed many related gene expressions like apoptosis and oxidative stress, cycle, junction, and signaling pathway. Taken together, AFB1 and AFM1 were found to affect cytotoxicity and related gene changes in BME cells. Notably, this study reported that 2 mg/L of AFB1 and AFM1 affected the expression of methylation-related genes, and ultimately altered the rate of m6A methylation in RNA. It may provide a potential direction for toxins to indirectly regulate gene expression by affecting RNA methylation modification. Our research provides some novel insights and data about AFB1 and AFM1 toxicity in BME cells.

The paradoxical effects of progesterone on the eggshell quality of laying hens.

This study demonstrates the effects of progesterone on eggshell quality and ultrastructure by injecting progesterone into laying hens 2 and 5 h post-oviposition, respectively. Progesterone injected 2 h post-oviposition (P4-2 h) improved eggshell quality with a significant decrease (P < 0.01) in the thickness of the mammillary layer and a significant increase (P < 0.01) in the thickness of the effective layer in the eggshell ultrastructure compared to the control. Progesterone injected 5 h post-oviposition (P4-5 h) damaged the eggshell quality by significantly reducing (P < 0.01) the effective layer thickness. Progesterone injected delayed obviously (P < 0.01) the following oviposition. Moreover, the concentrations of Thr, Cys, Leu, Lys, and His in the eggshell membranes were significantly higher (P < 0.05) in the P4-2 h treated hens whereas Val and Lys were significantly lower (P < 0.05) in P4-5 h treated hens compared to the control. Therefore, progesterone shows paradoxical effects on eggshell quality depending on the injection time-points post-oviposition, which could explain the contradictions in previous related reports. P4 injected affected the content of amino acids in eggshell membranes, especially lysine which contributed to eggshell quality. In addition, P4 injected 2 h after oviposition improved eggshell quality by promoting the premature fusion of mammillary knobs. This work contributed to a novel insight to understanding the mechanism of improving eggshell quality.

Individual-specific and shared representations during episodic memory encoding and retrieval.

Although human memories seem unique to each individual, they are shared to a great extent across individuals. Previous studies have examined, separately, subject-specific and cross-subject shared representations during memory encoding and retrieval, but how shared memories are formed from individually encoded representations is not clearly understood. Using a unique fMRI design involving memory encoding and retrieval, and representational similarity analysis to link representations from different individuals, brain regions, and processing stages, the current study revealed that distributed brain regions showed both subject-specific and shared neural representations during both memory encoding and retrieval. Furthermore, different brain regions showed stage-specific representational strength, with the visual cortex showing greater unique and shared representations during encoding, whereas the left angular gyrus showing greater unique and shared representations during retrieval. The neural representations during encoding were transformed during retrieval, as shown by smaller cross-subject encoding-retrieval similarity (ERS) than cross-subject similarity either during encoding or during retrieval. This cross-subject and cross-stage similarity was found both within and across regions, with strong pattern similarity between the encoded representation in VVC and the retrieved representation in the angular gyrus. Simulation analysis further suggested that these patterns could be achieved by incorporating stage-specific representational strength, and cross-region reinstatement from encoding to retrieval, but not by a common transformation from encoding to retrieval across subjects. Together, our results shed light on how memory representations are encoded and transformed to maintain individual characteristics and at the same time to create shared representations to facilitate interpersonal communication.

The differences of gonadal hormones and uterine transcriptome during shell calcification of hens laying hard or weak-shelled eggs.

BACKGROUND: Eggshell breaking strength is critical to reduce egg breaking rate and avoid economic loss. The process of eggshell calcification initiates with the egg entering the uterus and lasts about 18 h. It follows a temporal sequence corresponding to the initiation, growth and termination periods of shell calcification. During each period of shell calcification, our study investigated the differences of gonadal hormones and uterine transcriptome in laying hens producing a high or low breaking strength shell. RESULTS: 60 Hy-line Brown laying hens were selected and divided into two groups according to eggshell breaking strength. Eggshell breaking strength of 44.57 ± 0.91 N and 26.68 ± 0.38 N were considered to be the high strength group (HS) and low strength group (LS), respectively. The results showed that mammillary thickness and mammillary knob width of eggshells were significantly lower in the HS. Serum progesterone (P4) and 1,25-dihydroxy vitamin D3 [1,25-(OH)2D3] were significantly higher in the HS compared to the LS during the initiation period of calcification. Serum estradiol (E2) and calcium did not change significantly. All factors mentioned above had no significant differences in the growth and termination periods of calcification. The relative expression of CaBP-D28k and PMCA 1b were not significantly different between HS and LS. The relative expression of NCX1 was significantly higher in HS compared to LS. Moreover, 1777 differentially expressed genes (DEGs) were obtained in the initiation period of calcification. However, few DEGs were identified in the growth or termination periods of calcification. 30 DEGs were selected as candidate genes involved in eggshell calcification during the initiation period of calcification by the analysis of GO terms and KEGG pathways. CONCLUSIONS: Our study concluded that mammillary thickness and mammillary knob width of the HS were significantly lower than LS. P4 and 1,25-(OH)2D3 were significantly higher in the initiation period of HS. They may impact initial calcification when the mammillary layer is formed. The initiation period of calcification determined eggshell strength rather than the growth or termination periods. We inferred P4 or 1,25-(OH)2D3 may effect the ultrastructure of the mammillary layer by regulating the expression of uterine genes.

Disparity level identification using the voxel-wise Gabor model of fMRI data.

Perceiving disparities is the intuitive basis for our understanding of the physical world. Although many electrophysiology studies have revealed the disparity-tuning characteristics of the neurons in the visual areas of the macaque brain, neuron population responses to disparity processing have seldom been investigated. Many disparity studies using functional magnetic resonance imaging (fMRI) have revealed the disparity-selective visual areas in the human brain. However, it is unclear how to characterize neuron population disparity-tuning responses using fMRI technique. In the present study, we constructed three voxel-wise encoding Gabor models to predict the voxel responses to novel disparity levels and used a decoding method to identify the new disparity levels from population responses in the cortex. Among the three encoding models, the fine-coarse model (FCM) that used fine/coarse disparities to fit the voxel responses to disparities outperformed the single model and uncrossed-crossed model. Moreover, the FCM demonstrated high accuracy in predicting voxel responses in V3A complex and high accuracy in identifying novel disparities from responses in V3A complex. Our results suggest that the FCM can better characterize the voxel responses to disparities than the other two models and V3A complex is a critical visual area for representing disparity information.

Aberrant Connectivity in Mild Cognitive Impairment and Alzheimer Disease Revealed by Multimodal Neuroimaging Data.

BACKGROUND: Making use of multimodal data simultaneously to understand the neural mechanism of mild cognitive impairment (MCI) has been in the focus nowadays. The simultaneous use of multimodal data can take advantage of each modality which may only provide the view of one specific aspect of the brain. OBJECTIVE: To this end, the present study used structural magnetic resonance imaging (sMRI), fluorodeoxyglucose positron emission tomography (FDG-PET) and florbetapir PET to reveal the integrated brain network between MCI and normal controls (NCs). METHODS: In this study, 116 MCI, 116 NC and 116 Alzheimer disease (AD) subjects from the Alzheimer's Disease Neuroimaging Initiative were included for the evaluation of the brain covariance graphic model. Sparse inverse covariance estimation was utilized to get the graphic model. RESULTS: The connections among different brain regions were quite different between NC and MCI or between MCI and AD subjects (p < 0.01). The number of connections, which were represented by the covariance among different brain regions in the graphic model, decreased from NC to MCI and then AD, especially in the temporal lobe, occipital-parietal lobe and parietal-temporal lobe. CONCLUSION: These findings are good evidence to reveal the difference between MCI or AD and NC, and enhance the understanding of MCI.

Stereoscopic processing of crossed and uncrossed disparities in the human visual cortex.

BACKGROUND: Binocular disparity provides a powerful cue for depth perception in a stereoscopic environment. Despite increasing knowledge of the cortical areas that process disparity from neuroimaging studies, the neural mechanism underlying disparity sign processing [crossed disparity (CD)/uncrossed disparity (UD)] is still poorly understood. In the present study, functional magnetic resonance imaging (fMRI) was used to explore different neural features that are relevant to disparity-sign processing. METHODS: We performed an fMRI experiment on 27 right-handed healthy human volunteers by using both general linear model (GLM) and multi-voxel pattern analysis (MVPA) methods. First, GLM was used to determine the cortical areas that displayed different responses to different disparity signs. Second, MVPA was used to determine how the cortical areas discriminate different disparity signs. RESULTS: The GLM analysis results indicated that shapes with UD induced significantly stronger activity in the sub-region (LO) of the lateral occipital cortex (LOC) than those with CD. The results of MVPA based on region of interest indicated that areas V3d and V3A displayed higher accuracy in the discrimination of crossed and uncrossed disparities than LOC. The results of searchlight-based MVPA indicated that the dorsal visual cortex showed significantly higher prediction accuracy than the ventral visual cortex and the sub-region LO of LOC showed high accuracy in the discrimination of crossed and uncrossed disparities. CONCLUSIONS: The results may suggest the dorsal visual areas are more discriminative to the disparity signs than the ventral visual areas although they are not sensitive to the disparity sign processing. Moreover, the LO in the ventral visual cortex is relevant to the recognition of shapes with different disparity signs and discriminative to the disparity sign.

Structural covariance networks across healthy young adults and their consistency.

PURPOSE: To investigate structural covariance networks (SCNs) as measured by regional gray matter volumes with structural magnetic resonance imaging (MRI) from healthy young adults, and to examine their consistency and stability. MATERIALS AND METHODS: Two independent cohorts were included in this study: Group 1 (82 healthy subjects aged 18-28 years) and Group 2 (109 healthy subjects aged 20-28 years). Structural MRI data were acquired at 3.0T and 1.5T using a magnetization prepared rapid-acquisition gradient echo sequence for these two groups, respectively. We applied independent component analysis (ICA) to construct SCNs and further applied the spatial overlap ratio and correlation coefficient to evaluate the spatial consistency of the SCNs between these two datasets. RESULTS: Seven and six independent components were identified for Group 1 and Group 2, respectively. Moreover, six SCNs including the posterior default mode network, the visual and auditory networks consistently existed across the two datasets. The overlap ratios and correlation coefficients of the visual network reached the maximums of 72% and 0.71. CONCLUSION: This study demonstrates the existence of consistent SCNs corresponding to general functional networks. These structural covariance findings may provide insight into the underlying organizational principles of brain anatomy.

Motor Imagery Recognition Based on GMM-JCSFE Model.

Features from EEG microstate models, such as time-domain statistical features and state transition probabilities, are typically manually selected based on experience. However, traditional microstate models assume abrupt transitions between states, and the classification features can vary among individuals due to personal differences. To date, both empirical and theoretical classification results of EEG microstate features have not been entirely satisfactory. Here, we introduce an enhanced feature extraction method that combines Joint label-Common and label-Specific Feature Exploration (JCSFE) with Gaussian Mixture Models (GMM) to explore microstate features. First, GMMs are employed to represent the smooth transitions of EEG spatiotemporal features within microstate models. Second, category-common and category-specific features are identified by applying regularization constraints to linear classifiers. Third, a graph regularizer is used to extract subject-invariant microstate features. Experimental results on publicly available datasets demonstrate that the proposed model effectively encodes microstate features and improves the accuracy of motor imagery recognition across subjects. The primary code is accessible for download from the website: https://github.com/liaoliao3450/GMM-JCSFE.

s-TBN: A New Neural Decoding Model to Identify Stimulus Categories From Brain Activity Patterns.

Neural decoding is still a challenging and a hot topic in neurocomputing science. Recently, many studies have shown that brain network patterns containing rich spatiotemporal structural information represent the brain's activation information under external stimuli. In the traditional method, brain network features are directly obtained using the standard machine learning method and provide to a classifier, subsequently decoding external stimuli. However, this method cannot effectively extract the multidimensional structural information hidden in the brain network. Furthermore, studies on tensors have show that the tensor decomposition model can fully mine unique spatiotemporal structural characteristics of a spatiotemporal structure in data with a multidimensional structure. This research proposed a stimulus-constrained Tensor Brain Network (s-TBN) model that involves the tensor decomposition and stimulus category-constraint information. The model was verified on real neuroimaging data obtained via magnetoencephalograph and functional mangetic resonance imaging). Experimental results show that the s-TBN model achieve accuracy matrices of greater than 11.06% and 18.46% on the accuracy matrix compared with other methods on two modal datasets. These results prove the superiority of extracting discriminative characteristics using the STN model, especially for decoding object stimuli with semantic information.

A feature enhanced EEG compression model using asymmetric encoding-decoding network.

Objective.Recently, the demand for wearable devices using electroencephalography (EEG) has increased rapidly in many fields. Due to its volume and computation constraints, wearable devices usually compress and transmit EEG to external devices for analysis. However, current EEG compression algorithms are not tailor-made for wearable devices with limited computing and storage. Firstly, the huge amount of parameters makes it difficult to apply in wearable devices; secondly, it is tricky to learn EEG signals' distribution law due to the low signal-to-noise ratio, which leads to excessive reconstruction error and suboptimal compression performance.Approach.Here, a feature enhanced asymmetric encoding-decoding network is proposed. EEG is encoded with a lightweight model, and subsequently decoded with a multi-level feature fusion network by extracting the encoded features deeply and reconstructing the signal through a two-branch structure.Main results.On public EEG datasets, motor imagery and event-related potentials, experimental results show that the proposed method has achieved the state of the art compression performance. In addition, the neural representation analysis and the classification performance of the reconstructed EEG signals also show that our method tends to retain more task-related information as the compression ratio increases and retains reliable discriminative information after EEG compression.Significance.This paper tailors an asymmetric EEG compression method for wearable devices that achieves state-of-the-art compression performance in a lightweight manner, paving the way for the application of EEG-based wearable devices.

Decoding Human Interaction Type from Inter-brain Synchronization by Using EEG Brain Network.

Cooperation and competition are two common forms of interpersonal interactions and exploring inter-brain synchronization in these two forms can help to further deliberate the underlying neural mechanisms of interpersonal interactions. Recently, studies revealed that electrode-paired inter-brain synchronization plays an important role in human interactions. This study investigated the neural correlates of interpersonal synchronization at the brain network scale and interaction type. Firstly, the network-wise inter-brain synchronization (NIBS) index reflecting cross-brain network synchronization from the global brain perspective was advanced. Secondly, statistical analysis demonstrated that there are differences in NIBS activities between cooperative and competitive interactions. And a row-filtered depthwise separable convolution network was proposed to classify the NIBS features. Results of EEG hyper-scanning data showed significant differences in NIBS between cooperative and competitive tasks, and a comparative study manifested that the cross-brain synchronization in cooperative tasks is more consistent than that of competitive tasks. The neural decoder using a modified convolution network achieved a peak accuracy of 96.05% under the binary classification(cooperation vs competition).

Advanced Reinforcement Learning and Its Connections with Brain Neuroscience.

In recent years, brain science and neuroscience have greatly propelled the innovation of computer science. In particular, knowledge from the neurobiology and neuropsychology of the brain revolutionized the development of reinforcement learning (RL) by providing novel interpretable mechanisms of how the brain achieves intelligent and efficient decision making. Triggered by this, there has been a boom in research about advanced RL algorithms that are built upon the inspirations of brain neuroscience. In this work, to further strengthen the bidirectional link between the 2 communities and especially promote the research on modern RL technology, we provide a comprehensive survey of recent advances in the area of brain-inspired/related RL algorithms. We start with basis theories of RL, and present a concise introduction to brain neuroscience related to RL. Then, we classify these advanced RL methodologies into 3 categories according to different connections of the brain, i.e., micro-neural activity, macro-brain structure, and cognitive function. Each category is further surveyed by presenting several modern RL algorithms along with their mathematical models, correlations with the brain, and open issues. Finally, we introduce several important applications of RL algorithms, followed by the discussions of challenges and opportunities for future research.

Applying Interleaving Strategy of Learning Materials and Perceptual Modality to Address Secondary Students' Need to Restore Cognitive Capacity.

Online courses are prevalent around the world, especially during the COVID-19 pandemic. Long hours of highly demanding online learning can lead to mental fatigue and cognitive depletion. According to Attention Restoration Theory, 'being away' or a mental shift could be an important strategy to allow a person to recover from the cognitive overload. The present study aimed to test the interleaving strategy as a mental shift method to help sustain students' online learning attention and to improve learning outcomes. A total of 81 seventh-grade Chinese students were randomly assigned to four learning conditions: blocked (by subject matter) micro-lectures with auditory textual information (B-A condition), blocked (by subject matter) micro-lectures with visual textual information (B-V condition), interleaved (by subject matter) micro-lectures with auditory textual information (I-A condition), and interleaved micro-lectures by both perceptual modality and subject matter (I-all condition). We collected self-reported data on subjective cognitive load (SCL) and attention level, EEG data during the 40 min of online learning, and test results to assess learning outcomes. The results showed that the I-all condition showed the best overall outcomes (best performance, low SCL, and high attention). This study suggests that interleaving by both subject matter and perceptual modality should be preferred in scheduling and planning online classes.

Domain adaptation based on rough adjoint inconsistency and optimal transport for identifying autistic patients.

BACKGROUND AND OBJECTIVE: Computer aided diagnosis technology has been widely used to diagnose autism spectrum disorder (ASD) from neural images. The performance of the model usually depends largely on a sufficient number of training samples that reflect the real sample distribution. Due to the lack of labelled neural images data, multisite data are often pooled together to expand the sample size. However, the heterogeneity among sites will inevitably lead to a decline in the generalization of models. To solve this problem, we propose a multisource unsupervised domain adaptation method using rough adjoint inconsistency and optimal transport. METHODS: First, we define the concept of rough adjoint inconsistency and propose a double quantization method based on rough adjoint inconsistency and Dempster-Shafer (D-S) evidence theory to estimate the weight coefficient of each source domain to accurately describe the importance of each source domain to the target domain. Second, using optimal transport theory, we weaken the data distribution differences between domains and solve the problem of class imbalance by adjusting the sampling weights among classes. RESULTS: The ASD recognition accuracy of the proposed method is improved on all eight tasks, which are 70.67%, 64.86%, 62.50%, 70.80%, 73.08%, 71.19%, 75.41% and 75.76%, respectively. Our proposed model achieves superior performance compared to traditional machine learning methods and other recently proposed deep learning model. CONCLUSIONS: Our method demonstrates that the fusion of rough adjoint inconsistency and optimal transport can be a powerful tool for identifying ASD and quantifying the correlations between domains.

Categorizing objects from MEG signals using EEGNet.

Magnetoencephalography (MEG) signals have demonstrated their practical application to reading human minds. Current neural decoding studies have made great progress to build subject-wise decoding models to extract and discriminate the temporal/spatial features in neural signals. In this paper, we used a compact convolutional neural network-EEGNet-to build a common decoder across subjects, which deciphered the categories of objects (faces, tools, animals, and scenes) from MEG data. This study investigated the influence of the spatiotemporal structure of MEG on EEGNet's classification performance. Furthermore, the EEGNet replaced its convolution layers with two sets of parallel convolution structures to extract the spatial and temporal features simultaneously. Our results showed that the organization of MEG data fed into the EEGNet has an effect on EEGNet classification accuracy, and the parallel convolution structures in EEGNet are beneficial to extracting and fusing spatial and temporal MEG features. The classification accuracy demonstrated that the EEGNet succeeds in building the common decoder model across subjects, and outperforms several state-of-the-art feature fusing methods.

Rapidly Decoding Image Categories From MEG Data Using a Multivariate Short-Time FC Pattern Analysis Approach.

Recent advances in the development of multivariate analysis methods have led to the application of multivariate pattern analysis (MVPA) to investigate the interactions between brain regions using graph theory (functional connectivity, FC) and decode visual categories from functional magnetic resonance imaging (fMRI) data from a continuous multicategory paradigm. To estimate stable FC patterns from fMRI data, previous studies required long periods in the order of several minutes, in comparison to the human brain that categories visual stimuli within hundreds of milliseconds. Constructing short-time dynamic FC patterns in the order of milliseconds and decoding visual categories is a relatively novel concept. In this study, we developed a multivariate decoding algorithm based on FC patterns and applied it to magnetoencephalography (MEG) data. MEG data were recorded from participants presented with image stimuli in four categories (faces, scenes, animals and tools). MEG data from 17 participants demonstrate that short-time dynamic FC patterns yield brain activity patterns that can be used to decode visual categories with high accuracy. Our results show that FC patterns change over the time window, and FC patterns extracted in the time window of 0∼200 ms after the stimulus onset were most stable. Further, the categorizing accuracy peaked (the mean binary accuracy is above 78.6% at individual level) in the FC patterns estimated within the 0∼200 ms interval. These findings elucidate the underlying connectivity information during visual category processing on a relatively smaller time scale and demonstrate that the contribution of FC patterns to categorization fluctuates over time.

Domain Adaptation Using a Three-Way Decision Improves the Identification of Autism Patients from Multisite fMRI Data.

Machine learning methods are widely used in autism spectrum disorder (ASD) diagnosis. Due to the lack of labelled ASD data, multisite data are often pooled together to expand the sample size. However, the heterogeneity that exists among different sites leads to the degeneration of machine learning models. Herein, the three-way decision theory was introduced into unsupervised domain adaptation in the first time, and applied to optimize the pseudolabel of the target domain/site from functional magnetic resonance imaging (fMRI) features related to ASD patients. The experimental results using multisite fMRI data show that our method not only narrows the gap of the sample distribution among domains but is also superior to the state-of-the-art domain adaptation methods in ASD recognition. Specifically, the ASD recognition accuracy of the proposed method is improved on all the six tasks, by 70.80%, 75.41%, 69.91%, 72.13%, 71.01% and 68.85%, respectively, compared with the existing methods.

Mitigation Effects of Bentonite and Yeast Cell Wall Binders on AFB1, DON, and OTA Induced Changes in Laying Hen Performance, Egg Quality, and Health.

The objective of this study was to evaluate the efficacy of mycotoxin binders in reducing the adverse effects of co-occurring dietary aflatoxin B1 (AFB1), deoxynivalenol (DON) and ochratoxin A (OTA) on laying hens. Three hundred and sixty 26-week-old Roman laying hens were randomly allocated into four experimental groups with 10 replicates of nine birds each. The four groups received either a basal diet (BD; Control), a BD supplemented with 0.15 mg/kg AFB1 + 1.5 mg/kg DON + 0.12 mg/kg OTA (Toxins), a BD + Toxins with Toxo-HP binder (Toxins + HP), or a BD + Toxins with TOXO XL binder (Toxins + XL) for 12 weeks. Compared to the control, dietary supplementation of mycotoxins decreased (P < 0.10) total feed intake, total egg weight, and egg-laying rate, but increased feed/egg ratio by 2.5-6.1% and mortality during various experimental periods. These alterations induced by mycotoxins were alleviated by supplementation with both TOXO HP and XL binders (P < 0.10). Furthermore, dietary mycotoxins reduced (P < 0.05) eggshell strength by 12.3% and caused an accumulation of 249 µg/kg of DON in eggs at week 12, while dietary supplementation with TOXO HP or XL mitigated DON-induced changes on eggshell strength and prevented accumulation of DON in eggs (P < 0.05). Moreover, dietary mycotoxins increased relative liver weight, but decreased spleen and proventriculus relative weights by 11.6-22.4% (P < 0.05). Mycotoxin exposure also increased alanine aminotransferase activity and reduced immunoglobulin (Ig) A, IgM, and IgG concentrations in serum by 9.2-26.1% (P < 0.05). Additionally, mycotoxin exposure induced histopathological damage and reduced villus height, villus height/crypt depth, and crypt depth in duodenum, jejunum and (or) ileum (P < 0.05). Notably, most of these histological changes were mitigated by supplementation with both TOXO HP and XL (P < 0.05). In conclusion, the present study demonstrated that the mycotoxin binders TOXO HP and XL can help to mitigate the combined effects of AFB1, DON, and OTA on laying hen performance, egg quality, and health.