Closed-loop neuromodulation, especially using the phase of the electroencephalography (EEG) rhythm to assess the real-time brain state and optimize the brain stimulation process, is becoming a hot research topic. Because the EEG signal is non-stationary, the commonly used EEG phase-based prediction methods have large variances, which may reduce the accuracy of the phase prediction. In this study, we proposed a machine learning-based EEG phase prediction network, which we call EEG phase prediction network (EPN), to capture the overall rhythm distribution pattern of subjects and map the instantaneous phase directly from the narrow-band EEG data. We verified the performance of EPN on pre-recorded data, simulated EEG data, and a real-time experiment. Compared with widely used state-of-the-art models (optimized multi-layer filter architecture, auto-regress, and educated temporal prediction), EPN achieved the lowest variance and the greatest accuracy. Thus, the EPN model will provide broader applications for EEG phase-based closed-loop neuromodulation.
Humans ; Electroencephalography/methods* ; Brain/physiology* ; Machine Learning ; Signal Processing, Computer-Assisted ; Male ; Adult ; Neural Networks, Computer ; Brain Waves/physiology*

Schizophrenia (SZ) stands as a severe psychiatric disorder. This study applied diffusion tensor imaging (DTI) data in conjunction with graph neural networks to distinguish SZ patients from normal controls (NCs) and showcases the superior performance of a graph neural network integrating combined fractional anisotropy and fiber number brain network features, achieving an accuracy of 73.79% in distinguishing SZ patients from NCs. Beyond mere discrimination, our study delved deeper into the advantages of utilizing white matter brain network features for identifying SZ patients through interpretable model analysis and gene expression analysis. These analyses uncovered intricate interrelationships between brain imaging markers and genetic biomarkers, providing novel insights into the neuropathological basis of SZ. In summary, our findings underscore the potential of graph neural networks applied to multimodal DTI data for enhancing SZ detection through an integrated analysis of neuroimaging and genetic features.
Humans ; Schizophrenia/pathology* ; Diffusion Tensor Imaging/methods* ; Male ; Female ; Adult ; Brain/metabolism* ; Young Adult ; Middle Aged ; White Matter/pathology* ; Gene Expression ; Nerve Net/diagnostic imaging* ; Graph Neural Networks

Objective:To construct A novel scoring model of immune signatures for colon cancer based on machine learning,which improve the survival prediction and immune therapy.Methods:Screening immune signatures from 1 301 immune-related genes(IRG)by the combined strategy of Lasso+bootstrap+multi Cox to calculate IRG scores of colon cancer patients from TCGA databases,and comprehensive the differences on function,prognostic status and immune therapy between high IRG scores group and IRG scores group.Results:Groups based on IRG scores were significantly different on the prognostic status of colon cancer patients,which were validated by other independent datasets.The IRG scores also could assess the effect of immune therapy of colon cancer.Conclusion:This study provides ideas for immune therapy and researches of colon cancer based on immune genes,and IRG scores can be used to assess the prognosis of colon cancer patients.

Objective To observe the drainage pathways of interstitial fluid(ISF)and substance clearance pattern in rat brain with fluorescence tracing imaging and treacer-based MRI.Methods Thirty-three male SD rats were randomly divided into fluorescence tracing group(F group,n=18)and treacer-based MRI group(MRI group,n=15),then further divided into thalamic,hippocampal and caudate nucleus subgroups,respectively.Evans blue was injected to rats in F group,and cardiac perfusion was performed after injection,then brain tissue was harvested,and frozen sections were made to observe the drainage pathways of IFS in different subgroups.MRI was performed on rats in MRI group before and after injection of gadolinium-diethylenetriamine pentaacetic acid(Gd-DTPA)to observe signal intensity in ROI of brain regions in different subgroups,the signal unit ratio was calculated,and the changing trend was explored.Results ISF in thalamus,hippocampus and caudate nucleus had different dominant drainage pathways,and the time of tracer reached to adjacent brain regions and whole brain in F group were different.In MRI group,within 4 h after injection of Gd-DTPA,there were differences in direction and clearance rate among tracer in thalamus,hippocampus and caudate nucleus,mainly manifesting as the tracer in thalamus and hippocampus drained to the ipsilateral cortex and lateral ventricle,while the tracer in the caudate nucleus diffused to the cortex and midbrain,and there were differences of the peak time of tracer signal among adjacent drainage brain regions.Conclusion Fluorescence and MR dual-mode imaging showed that there were differences in the dominant drainage pathways of IFS and clearance rates of small molecule substances among hypothalamus,hippocampus and caudate nucleus of rats.

Objective To observe the alterations in functional complexity of brain regions in autism spectrum disorder(ASD)patients and correlations with the predicted brain age.Methods Open brain resting-state functional MRI(rs-MRI)data of 93 ASD patients and 96 typically developing adolescents(healthy subjects)were downloaded.The functional complexity in brain regions were extracted with self-developed virtual digital brain software,and the alterations in functional complexity of brain regions in ASD patients and correlations with their ages were analyzed.Two networks were prospectively trained with data of 65 ASD patients and 67 healthy subjects as the training set to predict brain age,and the results were evaluated,and the predicting errors were compared using test set,i.e.the other 28 ASD patients and 29 healthy subjects.Results Compared to healthy subjects,on the basis of anatomical automatic labeling(AAL)atlas,ASD patients exhibited significantly reduced functional complexity based on Shannon entropy in the left precuneus,left cuneus and right parahippocampal gyrus.Conversely,functional complexity of ASD patients based on permutation entropy significantly increased in the left cuneus and right cerebellar Crus Ⅱ region.The left hippocampus showed reduced functional complexity based on Pearson correlation coefficient,while the left middle temporal gyrus showed increased functional complexity based on Pearson correlation coefficient.The functional complexity in brain regions of ASD patients were not closely correlated with ages(all|r|＜0.4).According to the trained fully connected network,the predicted brain ages of ASD patients and healthy subjects in test set were all lower than their physiological ages,but no significant difference was found between the prediction errors of ASD patients and healthy subjects(P=0.283).Conclusion Functional complexity changed in some brain region functions in ASD patients.The predicted brain ages of ASD patients based on the obtained fully connected network were on the low side,but not obviously affected by the alterations of functional complexity in brain regions.

Alzheimer's disease (AD) is associated with the impairment of white matter (WM) tracts. The current study aimed to verify the utility of WM as the neuroimaging marker of AD with multisite diffusion tensor imaging datasets [321 patients with AD, 265 patients with mild cognitive impairment (MCI), 279 normal controls (NC)], a unified pipeline, and independent site cross-validation. Automated fiber quantification was used to extract diffusion profiles along tracts. Random-effects meta-analyses showed a reproducible degeneration pattern in which fractional anisotropy significantly decreased in the AD and MCI groups compared with NC. Machine learning models using tract-based features showed good generalizability among independent site cross-validation. The diffusion metrics of the altered regions and the AD probability predicted by the models were highly correlated with cognitive ability in the AD and MCI groups. We highlighted the reproducibility and generalizability of the degeneration pattern of WM tracts in AD.
Humans ; White Matter/diagnostic imaging* ; Diffusion Tensor Imaging/methods* ; Alzheimer Disease/complications* ; Reproducibility of Results ; Cognition ; Cognitive Dysfunction/complications* ; Brain/diagnostic imaging*

Gray Matter/diagnostic imaging* ; Cerebral Cortex ; Brain ; Magnetic Resonance Imaging

Objective To investigate the effect of Sema6D knockdown on the proliferation, migration, invasion and angiogenesis-promoting ability of human osteosarcoma cell lines. Methods The expression of Sema6D in clinical tissues and cell lines of human osteosarcoma was detected. After the targeted siRNA transfection, the changes of proliferation, migration and invasion were measured by CCK-8, wound healing and Transwell experiments. HUVECs were co-cultured with tumor conditioned medium to detect their tube formation ability. And the expression of signal pathway proteins was detected by Western blot. Results Sema6D was highly expressed in human osteosarcoma tissues and cell lines(P < 0.05). After silencing Sema6D, the proliferation, migration and invasion of 143B and MG63 cells decreased significantly(P < 0.05), the angiogenesis ability of HUVECs decreased in vitro(P < 0.01), and the expression of PI3K/AKT/mTOR and ERK-related signal pathway proteins decreased(P < 0.05). Conclusion Sema6D is overexpressed in human osteosarcoma tissues and cell lines. Knockdown of Sema6D expression level could inhibit the proliferation, migration, invasion and angiogenesis-promoting ability of human osteosarcoma cells via reducing PI3K/AKT/mTOR and ERK signal pathway.