An intelligent workflow for sub-nanoscale 3D reconstruction of intact synapses from serial section electron tomography.

BACKGROUND: As an extension of electron tomography (ET), serial section electron tomography (serial section ET) aims to align the tomographic images of multiple thick tissue sections together, to break through the volume limitation of the single section and preserve the sub-nanoscale voxel size. It could be applied to reconstruct the intact synapse, which expands about one micrometer and contains nanoscale vesicles. However, there are several drawbacks of the existing serial section ET methods. First, locating and imaging regions of interest (ROIs) in serial sections during the shooting process is time-consuming. Second, the alignment of ET volumes is difficult due to the missing information caused by section cutting and imaging. Here we report a workflow to simplify the acquisition of ROIs in serial sections, automatically align the volume of serial section ET, and semi-automatically reconstruct the target synaptic structure. RESULTS: We propose an intelligent workflow to reconstruct the intact synapse with sub-nanometer voxel size. Our workflow includes rapid localization of ROIs in serial sections, automatic alignment, restoration, assembly of serial ET volumes, and semi-automatic target structure segmentation. For the localization and acquisition of ROIs in serial sections, we use affine transformations to calculate their approximate position based on their relative location in orderly placed sections. For the alignment of consecutive ET volumes with significantly distinct appearances, we use multi-scale image feature matching and the elastic with belief propagation (BP-Elastic) algorithm to align them from coarse to fine. For the restoration of the missing information in ET, we first estimate the number of lost images based on the pixel changes of adjacent volumes after alignment. Then, we present a missing information generation network that is appropriate for small-sample of ET volume using pre-training interpolation network and distillation learning. And we use it to generate the missing information to achieve the whole volume reconstruction. For the reconstruction of synaptic ultrastructures, we use a 3D neural network to obtain them quickly. In summary, our workflow can quickly locate and acquire ROIs in serial sections, automatically align, restore, assemble serial sections, and obtain the complete segmentation result of the target structure with minimal manual manipulation. Multiple intact synapses in wild-type rat were reconstructed at a voxel size of 0.664 nm/voxel to demonstrate the effectiveness of our workflow. CONCLUSIONS: Our workflow contributes to obtaining intact synaptic structures at the sub-nanometer scale through serial section ET, which contains rapid ROI locating, automatic alignment, volume reconstruction, and semi-automatic synapse reconstruction. We have open-sourced the relevant code in our workflow, so it is easy to apply it to other labs and obtain complete 3D ultrastructures which size is similar to intact synapses with sub-nanometer voxel size.

Joint reconstruction of neuron and ultrastructure via connectivity consensus in electron microscope volumes.

BACKGROUND: Nanoscale connectomics, which aims to map the fine connections between neurons with synaptic-level detail, has attracted increasing attention in recent years. Currently, the automated reconstruction algorithms in electron microscope volumes are in great demand. Most existing reconstruction methodologies for cellular and subcellular structures are independent, and exploring the inter-relationships between structures will contribute to image analysis. The primary goal of this research is to construct a joint optimization framework to improve the accuracy and efficiency of neural structure reconstruction algorithms. RESULTS: In this investigation, we introduce the concept of connectivity consensus between cellular and subcellular structures based on biological domain knowledge for neural structure agglomeration problems. We propose a joint graph partitioning model for solving ultrastructural and neuronal connections to overcome the limitations of connectivity cues at different levels. The advantage of the optimization model is the simultaneous reconstruction of multiple structures in one optimization step. The experimental results on several public datasets demonstrate that the joint optimization model outperforms existing hierarchical agglomeration algorithms. CONCLUSIONS: We present a joint optimization model by connectivity consensus to solve the neural structure agglomeration problem and demonstrate its superiority to existing methods. The intention of introducing connectivity consensus between different structures is to build a suitable optimization model that makes the reconstruction goals more consistent with biological plausible and domain knowledge. This idea can inspire other researchers to optimize existing reconstruction algorithms and other areas of biological data analysis.

An innovative method for screening and evaluating the degree of diabetic retinopathy and drug treatment based on artificial intelligence algorithms.

Current methods of evaluating the degree of diabetic retinopathy are highly subjective and have no quantitative standard. To objectively evaluate the slight changes in tissue structure during the early stage of retinal diseases, a subjective interpretation and qualitative analysis of the pathological sections of retinal HE in diabetic animals is required for screening and evaluating the degree of diabetic retinopathy and drug efficacy. To develop an innovative method for screening and evaluating the degree of diabetic retinopathy and drug treatment based on artificial intelligence algorithms. Based on the change law of the early nerve fiber layer and the ganglion cells, we get disparate characteristics of the microscopic image of diabetes animal retina HE slices. Using image recognition and deep learning methods on these HE slices, we can identify the changes in the ganglion cells and nerve fiber layer for diagnosing early retinopathy and evaluated the therapeutic effect of the potential drugs. We conduct quantitative calculation per unit length of the nerve fiber layer and total area of the nerve fiber layer to identify biology significance of edema. Additionally, we also perform quantitative calculation with the number of unit area ganglion cells to identify the section in biology cell hyperplasia. Finally, we get the significance of quantitative calculation on the unit cell area to identify ganglion cell shriveling in biology. In addition to the evaluation of the disease degree and changes, we also obtained retinal HE sections after different drug interventions and evaluated the therapeutic effect of the drugs. This study presents a novel quantitative method for screening and evaluating of diabetic retinopathy and drug efficacy.

Oseltamivir hydroxamate and acyl sulfonamide derivatives as influenza neuraminidase inhibitors.

Tamiflu, the ethyl ester form of oseltamivir carboxylic acid (OC), is the first orally available anti-influenza drug for the front-line therapeutic option. In this study, the OC-hydroxamates, OC-sulfonamides and their guanidino congeners (GOC) were synthesized. Among them, an OC-hydroxamate 7d bearing an O-(2-indolyl)propyl substituent showed potent NA inhibition (IC50 = 6.4 nM) and good anti-influenza activity (EC50 = 60.1 nM) against the wild-type H1N1 virus. Two GOC-hydroxamates (9b and 9d) and one GOC-sulfonamide (12a) were active to the tamiflu-resistant H275Y virus (EC50 = 2.3-6.9 µM).

Multi-consensus formation control by artificial potential field based on velocity threshold.

This study proposes a multi-consensus formation control algorithm by artificial potential field (APF) method based on velocity threshold. The algorithm improves the multi-consensus technique. This algorithm can split a group of agents into multiple agent groups. Note that the algorithm can easily complete the queue transformation as long as the entire proxy group is connected initially and no specific edges need to be removed. Furthermore, collision avoidance and maintenance of existing communication connectivity should be considered during the movement of all agents. Therefore, we design a new swarm motion potential function. The stability of multi-consensus formation control has proven to be effective in avoiding collisions, maintaining connectivity, and generating formations. The final numerical simulation results show the role of the controller we designed.

DeepMulticut: Deep Learning of Multicut Problem for Neuron Segmentation from Electron Microscopy Volume.

Superpixel aggregation is a powerful tool for automated neuron segmentation from electron microscopy (EM) volume. However, existing graph partitioning methods for superpixel aggregation still involve two separate stages-model estimation and model solving, and therefore model error is inherent. To address this issue, we integrate the two stages and propose an end-to-end aggregation framework based on deep learning of the minimum cost multicut problem called DeepMulticut. The core challenge lies in differentiating the NPhard multicut problem, whose constraint number is exponential in the problem size. With this in mind, we resort to relaxing the combinatorial solver-the greedy additive edge contraction (GAEC)-to a continuous Soft-GAEC algorithm, whose limit is shown to be the vanilla GAEC. Such relaxation thus allows the DeepMulticut to integrate edge cost estimators, Edge-CNNs, into a differentiable multicut optimization system and allows a decision-oriented loss to feed decision quality back to the Edge-CNNs for adaptive discriminative feature learning. Hence, the model estimators, Edge-CNNs, can be trained to improve partitioning decisions directly while beyond the NP-hardness. Also, we explain the rationale behind the DeepMulticut framework from the perspective of bi-level optimization. Extensive experiments on three public EM datasets demonstrate the effectiveness of the proposed DeepMulticut.

Internet addiction in adolescent psychiatric patient population: A hospital-based Study from China.

BACKGROUND: Internet is gradually reshaping adolescents' behaviors in China. It is important to identify the prevalence and risk factors to predict Internet addiction among adolescent psychiatric patients. METHODS: The survey was conducted among inpatient or outpatient adolescent patients with mental disorders. All participants were interviewed face-to-face and completed Young Internet Addiction Test and other relevant assessments. Binary logistic regression analysis was used to exam between-group differences of internet addiction. RESULT: The prevalence for internet addiction (mild to severe) in adolescent psychiatric patients was 80.2%, where the prevalence for "medium and severe internet addiction was 25.5%. Logistic regression analysis identified two independent predictors for "medium and severe internet addiction" including the total score of Adolescent Non-suicidal-self-injury (NSSI) Behavior Function Assessment Scale (ANBFAS) and the diagnosis of generalized anxiety disorder (GAD) (R2 =0.27, p = .02 for total ANBFAS score, p = .01 for GAD) in psychiatric adolescent patients. CONCLUSION: The prevalence of internet addiction is notably high among adolescent psychiatric patients in China. Coping for Internet addiction in adolescent psychiatric patients should conclude the coping of NSSI and GAD.

Validation of the developmental, dimensional and diagnostic interview -short version (3Di-sv) for diagnosis of autism spectrum disorder in Chinese children.

The purpose of this study was to evaluate the psychometric features of the Chinese version of the developmental dimensional diagnostic interview-short version (3Di-sv). A total sample of 138 children including 79 children with autism spectrum disorder (ASD) and 59 typically developing children completed the 3Di-sv interview. The Chinese version of the 3Di-sv has a good internal consistency (0.94). Test-retest analysis confirmed the instrument's time stability (0.89). The instrument's concurrent validity with the Autism Behavior Checklist (ABC), the Childhood Autism Rating Scale (CARS) and clinical diagnosis was verified; the correlation between total scores was 0.72, 0.82 and 0.90, respectively. The 3Di-sv significantly distinguished between autistic children and non-autistic children in every area of autism symptoms. Optimal cutoffs were derived using receiver operating characteristics curves. Using clinical diagnosis as criterion, overall sensitivity was 98 % and specificity was 90 %. The study determined that the Chinese version of 3Di-sv can well distinguish autistic children from typically developing children.

Deep residual contextual and subpixel convolution network for automated neuronal structure segmentation in micro-connectomics.

BACKGROUND AND OBJECTIVE: The goal of micro-connectomics research is to reconstruct the connectome and elucidate the mechanisms and functions of the nervous system via electron microscopy (EM). Due to the enormous variety of neuronal structures, neuron segmentation is among most difficult tasks in connectome reconstruction, and neuroanatomists desperately need a reliable neuronal structure segmentation method to reduce the burden of manual labeling and validation. METHODS: In this article, we proposed an effective deep learning method based on a deep residual contextual and subpixel convolution network to obtain the neuronal structure segmentation in anisotropic EM image stacks. Furthermore, lifted multicut is used for post-processing to optimize the prediction and obtain the reconstruction results. RESULTS: On the ISBI EM segmentation challenge, the proposed method ranks among the top of the leader board and yields a Rand score of 0.98788. On the public data set of mouse piriform cortex, it achieves a Rand score of 0.9562 and 0.9318 in the different testing stacks. The evaluation scores of our method are significantly improved when compared with those of state-of-the-art methods. CONCLUSIONS: The proposed automatic method contributes to the development of micro-connectomics, which improves the accuracy of neuronal structure segmentation and provides neuroanatomists with an effective approach to obtain the segmentation and reconstruction of neurons.

Fear memory-associated synaptic and mitochondrial changes revealed by deep learning-based processing of electron microscopy data.

Serial section electron microscopy (ssEM) can provide comprehensive 3D ultrastructural information of the brain with exceptional computational cost. Targeted reconstruction of subcellular structures from ssEM datasets is less computationally demanding but still highly informative. We thus developed a region-CNN-based deep learning method to identify, segment, and reconstruct synapses and mitochondria to explore the structural plasticity of synapses and mitochondria in the auditory cortex of mice subjected to fear conditioning. Upon reconstructing over 135,000 mitochondria and 160,000 synapses, we find that fear conditioning significantly increases the number of mitochondria but decreases their size and promotes formation of multi-contact synapses, comprising a single axonal bouton and multiple postsynaptic sites from different dendrites. Modeling indicates that such multi-contact configuration increases the information storage capacity of new synapses by over 50%. With high accuracy and speed in reconstruction, our method yields structural and functional insight into cellular plasticity associated with fear learning.

Low mental health literacy and its association with depression, anxiety and poor sleep quality in Chinese elderly.

BACKGROUND: Mental health literacy (MHL) is rarely reported in the Chinese elderly. This study explored the pattern of MHL in the Chinese elderly in relation to depression, anxiety and poor sleep quality. METHODS: A cross-sectional study was conducted among older adults in Guangzhou, south China. Participants were investigated face-to-face using the Chinese National Mental Health Literacy Scale, the Patient Health Questionnaire-9 item (PHQ-9), the 7-item Generalized Anxiety Disorder (GAD-7) and the Pittsburgh Sleep Quality Index (PSQI). Mental Health Literacy Scale contains three dimensions: mental health knowledge, mental health skills (such as social support, cognitive reappraisal and attentional distraction) and mental health awareness. Multivariate logistic regression was used for examining the association between MHL and mental health. RESULTS: A total of 506 older adults were recruited. The percentage of depression, anxiety, and poor sleep quality were 16.6%, 7.9% and 40.9%, respectively. MHL dimensions independently associated with depression included cognitive reappraisal (OR = 1.95, p < .001), attentional distraction (OR = 0.61, p = 0.044) and awareness (OR = 0.56, p = 0.027). MHL dimensions independently associated with anxiety symptoms included cognitive reappraisal (OR = 1.90, p = 0.011) and attentional distraction (OR = 0.44, p = 0.016). MHL dimensions independently associated with poor sleep quality included social support (OR = 0.75, p = 0.022), cognitive reappraisal (OR = 1.55, p = 0.003) and attentional distraction (OR = 0.65, p = 0.016). CONCLUSION: Given the low MHL and its association with poor mental health in the Chinese elderly, policymakers and health professionals should improve the older adults' MHL, which could be conducive to the prevention and control of their mental health problems.

Automatic Reconstruction of Mitochondria and Endoplasmic Reticulum in Electron Microscopy Volumes by Deep Learning.

Together, mitochondria and the endoplasmic reticulum (ER) occupy more than 20% of a cell's volume, and morphological abnormality may lead to cellular function disorders. With the rapid development of large-scale electron microscopy (EM), manual contouring and three-dimensional (3D) reconstruction of these organelles has previously been accomplished in biological studies. However, manual segmentation of mitochondria and ER from EM images is time consuming and thus unable to meet the demands of large data analysis. Here, we propose an automated pipeline for mitochondrial and ER reconstruction, including the mitochondrial and ER contact sites (MAMs). We propose a novel recurrent neural network to detect and segment mitochondria and a fully residual convolutional network to reconstruct the ER. Based on the sparse distribution of synapses, we use mitochondrial context information to rectify the local misleading results and obtain 3D mitochondrial reconstructions. The experimental results demonstrate that the proposed method achieves state-of-the-art performance.

Boronate, trifluoroborate, sulfone, sulfinate and sulfonate congeners of oseltamivir carboxylic acid: Synthesis and anti-influenza activity.

Tamiflu readily undergoes endogenous hydrolysis to give oseltamivir carboxylic acid (OC) as the active anti-influenza agent to inhibit the viral neuraminidase (NA). GOC is derived from OC by replacing the 5-amino group with a guanidino group. In this study, OC and GOC congeners with the carboxylic acid bioisosteres of boronic acid, trifluoroborate, sulfone, sulfinic acid, sulfonic acid and sulfonate ester were first synthesized, starting with conversion of OC to a Barton ester, followed by halodecarboxylation to give the iodocyclohexene, which served as a pivotal intermediate for palladium-catalyzed coupling reactions with appropriate diboron and thiol reagents. The enzymatic and cell-based assays indicated that the GOC congeners consistently displayed better NA inhibition and anti-influenza activity than the corresponding OC congeners. The GOC sulfonic acid congener (7a) was the most potent anti-influenza agent, showing EC50 = 2.2 nM against the wild-type H1N1 virus, presumably because the sulfonic acid 7a was more lipophilic than GOC and exerted stronger interactions on the three arginine residues (R118, R292 and R371) in the NA active site. Although the trifluoroborates, sulfones and sulfonate esters did not have acidic proton, they still exhibited appreciable NA inhibitory activity, indicating that the polarized B-F and S→O bonds still made sufficient interactions with the tri-arginine motif.

Automatic Classification for the Type of Multiple Synapse Based on Deep Learning.

Recent studies have shown that the synaptic plasticity induced by development and learning can promote the formation of multiple synapse. With the rapid development of electron microscopy (EM) technology, we can closely observe the multiple synapse structure with high resolution. Although the multiple synapse has been widely researched by recent researchers, the classification accuracy for the type of multiple synapse has not been documented. In this paper, we propose an effective automatic classification method for the type of multiple synapse. The main steps are summarized as three parts: synaptic cleft segmentation, vesicle band segmentation, multiple synapse classification. The experiments on four datasets demonstrate that the proposed method can reach an average accuracy about 97%.

Automatic Detection and Segmentation of Mitochondria from SEM Images using Deep Neural Network.

Investigating the link between mitochondrial function and its physical structure is a hot topic in neurobiology research. With the rapid development of Scanning Electron Microscope (SEM), we can look closely into the fine mitochondrial structure with high resolution. Consequently, many meaningful researches have focused on how to detect and segment the mitochondria from EM images. Due to the complex background, hand-crafted features designed by traditional algorithms cannot provide satisfying results. In this paper, we propose an effective deep neural network improved from Mask R-CNN to produce the detection and segmentation results. On this base, we use the morphological processing and mitochondrial context information to rectify the local misleading results. The valuation was performed on two widely used datasets (FIB-SEM and ATUMSEM), and the results demonstrate that the proposed method has comparable performance than state-of-the-art methods.