RESUMEN
The visual systems of humans and nonhuman primates share many similarities in both anatomical and functional organization. Understanding the homology and differences between the two systems can provide important insights into the neural basis of visual perception and cognition. This research aims to investigate the homology between human and macaque visual systems based on connectivity, using diffusion tensor imaging and resting-state functional magnetic resonance imaging to construct structural and functional connectivity fingerprints of the visual systems in humans and macaques, and quantitatively analyze the connectivity patterns. By integrating multimodal magnetic resonance imaging, this research explored the homology and differences between the two systems. The results showed that 9 brain regions in the macaque visual system formed highly homologous mapping relationships with 11 brain regions in the human visual system, and the related brain regions between the two species showed highly structure homologous, with their functional organization being essentially conserved across species. Finally, this research generated a homology information map of the visual system for humans and macaques, providing a new perspective for subsequent cross-species analysis.
RESUMEN
Brain extraction (skull stripping) is an essential step in the magnetic resonance imaging (MRI) analysis of brain sciences. However, most of the current brain extraction methods that achieve satisfactory results for human brains are often challenged by non-human primate brains. Due to the small sample characteristics and the nature of thick-slice scanning of macaque MRI data, traditional deep convolutional neural networks (DCNNs) are unable to obtain excellent results. To overcome this challenge, this study proposed a symmetrical end-to-end trainable hybrid convolutional neural network (HC-Net). It makes full use of the spatial information between adjacent slices of the MRI image sequence and combines three consecutive slices from three axes for 3D convolutions, which reduces the calculation consumption and promotes accuracy. The HC-Net consists of encoding and decoding structures of 3D convolutions and 2D convolutions in series. The effective use of 2D convolutions and 3D convolutions relieves the underfitting of 2D convolutions to spatial features and the overfitting of 3D convolutions to small samples. After evaluating macaque brain data from different sites, the results showed that HC-Net performed better in inference time (approximately 13 s per volume) and accuracy (mean Dice coefficient reached 95.46%). The HC-Net model also had good generalization ability and stability in different modes of brain extraction tasks.
RESUMEN
Using the animal brain as a cross-species tool for human brain research based on imaging features can provide more potential to reveal comprehensive human brain analysis. Previous studies have shown that human Brodmann area 5 (BA5) and macaque PE are homologous regions. They are both involved in processes depth and direction information during the touch process in the arm movement. However, recent studies show that both BA5 and PE are not homogeneous. According to the cytoarchitecture, BA5 is subdivided into three different subregions, and PE can be subdivided into PEl, PEla, and PEm. The species homologous relationship among the subregions is not clear between BA5 and PE. At the same time, the subdivision of PE based on the anatomical connection of white matter fiber bundles needs more verification. This research subdivided the PE of macaques based on the anatomical connection of white matter fiber bundles. Two PE subregions are defined based on probabilistic fiber tracking, one on the anterior side and the other on the dorsal side. Finally, the research draws connectivity fingerprints with predefined homologous target areas for the BA5 and PE subregions to reveal the characteristics of structure and functions and gives the homologous correspondence identified.
RESUMEN
Synthesis of NiHCCo precursors via simple co-precipitation and nickel-cobalt tetraselenide composites grown on nitrogen-doped reduced graphene oxide (NiCoSe4/N-rGO) were fabricated using solvothermal method. The introduction of N-rGO used as a template effectively prevented agglomeration of NiCoSe4nanoparticles and provided more active sites, which greatly increased the electrochemical and electrical conductivity for NiCoSe4/N-rGO. NiCoSe4/N-rGO-20 presents a remarkably elevated specific capacity of 120 mA h g-1under current density of 1 A g-1. NiCoSe4/N-rGO-20 demonstrates an excellent cycle life and achieves a remarkable 83% retention rate over 3000 cycles with 10 A g-1. NiCoSe4/N-rGO-20//N-rGO asymmetric supercapacitor was constructed based on the NiCoSe4/N-rGO-20 as an anode, N-rGO as cathode by using 2 mol l-1KOH as an electrolyte. NiCoSe4/N-rGO-20//N-rGO ASC demonstrates an ultra-big energy density of 14 Wh kg-1and good circulation stability in the power density of 902 W kg-1. It is doubled in comparison to the NiCoSe4/N-rGO-20//rGO asymmetric supercapacitor (7 Wh kg-1). The NiCoSe4/N-rGO-20//N-rGO ASC capacity retention is still up to 93% over 5000 cycles (5 A g-1). The results reveal that this device would be a prospective cathode material of supercapacitors in actual applications.
RESUMEN
Research has shown that abnormal brain networks in patients with schizophrenia appear at different frequencies, but the relationship between these different frequencies is unclear. Therefore, it is necessary to use a multilayer network model to evaluate the integration of information from different frequency bands. To explore the mechanism of integration and separation in the multilayer network of schizophrenia, we constructed multilayer frequency brain network models in 50 patients with schizophrenia and 69 healthy subjects, and the entropy of the multiplex degree (EMD) and multilayer clustering coefficient (MCC) were calculated. The results showed that the ability to integrate and separate information in the multilayer network of patients was significantly higher than that of normal people. This difference was mainly reflected in the default mode network, sensorimotor network, subcortical network, and visual network. Among them, the subcortical network was different in both MCC and EMD outcomes. Furthermore, differences were found in the posterior cingulate gyrus, hippocampus, amygdala, putamen, pallidum, and thalamus. The thalamus and posterior cingulate gyrus were associated with the patient's symptom scores. Our results showed that the cross-frequency interaction ability of patients with schizophrenia was significantly enhanced, among which the subcortical network was the most active. This interaction may serve as a compensation mechanism for intralayer dysfunction.
RESUMEN
Accurately extracting brain tissue is a critical and primary step in brain neuroimaging research. Due to the differences in brain size and structure between humans and nonhuman primates, the performance of the existing tools for brain tissue extraction, working on macaque brain MRI, is constrained. A new transfer learning training strategy was utilized to address the limitations, such as insufficient training data and unsatisfactory model generalization ability, when deep neural networks processing the limited samples of macaque magnetic resonance imaging(MRI). First, the project combines two human brain MRI data modes to pre-train the neural network, in order to achieve faster training and more accurate brain extraction. Then, a residual network structure in the U-Net model was added, in order to propose a ResTLU-Net model that aims to improve the generalization ability of multiple research sites data. The results demonstrated that the ResTLU-Net, combined with the proposed transfer learning strategy, achieved comparable accuracy for the macaque brain MRI extraction tasks on different macaque brain MRI volumes that were produced by various medical centers. The mean Dice of the ResTLU-Net was 95.81% (no need for denoise and recorrect), and the method required only approximately 30-60 s for one extraction task on an NVIDIA 1660S GPU.
RESUMEN
Electroencephalograph (EEG) signals and graph theory measures have been widely used to characterize the brain functional networks of healthy individuals and patients by calculating the correlations between different electrodes over an entire time series. Although EEG signals have a high temporal resolution and can provide relatively stable results, the process of constructing and analyzing brain functional networks is inevitably complicated by high time complexity. Our goal in this research was to distinguish the brain function networks of schizophrenia patients from those of healthy participants during working memory tasks. Consequently, we utilized a method involving microstates, which are each characterized by a unique topography of electric potentials over an entire channel array, to reduce the dimension of the EEG signals during working memory tasks and then compared and analyzed the brain functional networks using the microstates time series (MTS) and original time series (OTS) of the schizophrenia patients and healthy individuals. We found that the right frontal and parietal-occipital regions neurons of the schizophrenia patients were less active than those of the healthy participants during working memory tasks. Notably, compared with OTS, the time needed to construct the brain functional networks was significantly reduced by using MTS. In conclusion, our results show that, like OTS, MTS can well distinguish the brain functional network of schizophrenia patients from those of healthy individuals during working memory tasks while greatly decreasing time complexity. MTS can thus provide a method for characterizing the original time series for the construction and analysis of EEG brain functional networks.
