A micropeptide TREMP encoded by lincR-PPP2R5C promotes Th2 cell differentiation by interacting with PYCR1 in allergic airway inflammation.

BACKGROUND: Allergic asthma is largely dominated by Th2 lymphocytes. Micropeptides in Th2 cells and asthma remain unmasked. Here, we aimed to demonstrate a micropeptide, T-cell regulatory micropeptide (TREMP), in Th2 cell differentiation in asthma. METHODS: TREMP translated from lincR-PPP2R5C was validated using Western blotting and mass spectrometry. TREMP knockout mice were generated using CRISPR/Cas9. Coimmunoprecipitation revealed that TREMP targeted pyrroline-5-carboxylate reductase 1 (PYCR1), which was further explored in vitro and in vivo. The levels of TREMP and PYCR1 in Th2 cells from clinical samples were determined by flow cytometry. RESULTS: TREMP, encoded by lincR-PPP2R5C, was in the mitochondrion. The lentivirus encoding TREMP promoted Th2 cell differentiation. In contrast, Th2 differentiation was suppressed in TREMP-/- CD4+ T cells. In the HDM-induced model of allergic airway inflammation, TREMP was increased in pulmonary tissues. Allergic airway inflammation was relieved in TREMP-/- mice treated with HDM. Mechanistically, TREMP interacted with PYCR1, which regulated Th2 differentiation via glycolysis. Glycolysis was decreased in Th2 cells from TREMP-/- mice and PYCR1-/- mice. Similar to TREMP-/- mice, allergic airway inflammation was mitigated in HDM-challenged PYCR1-/- mice. Moreover, we measured TREMP and PYCR1 in asthma patients. And we found that, compared with those in healthy controls, the levels of TREMP and PYCR1 in Th2 cells were significantly increased in asthmatic patients. CONCLUSIONS: The micropeptide TREMP encoded by lincR-PPP2R5C promoted Th2 differentiation in allergic airway inflammation by interacting with PYCR1 and enhancing glycolysis. Our findings highlight the importance of neglected micropeptides from noncoding RNAs in allergic diseases.

Fine particulate matter (PM2.5) induces inhibitory memory alveolar macrophages through the AhR/IL-33 pathway.

Fine particulate matter (PM2.5) concentrations have decreased in the past decade. The adverse effects of acute PM2.5 exposure on respiratory diseases have been well recognized. To explore the long-term effects of PM2.5 exposure on chronic obstructive pulmonary disease (COPD), mice were exposed to PM2.5 for 7 days and rest for 21 days, followed by challenges with lipopolysaccharide (LPS) and porcine pancreatic elastase (PPE). Unexpectedly, PM2.5 exposure and rest alleviated the disease severity and airway inflammatory responses in COPD-like mice. Although acute PM2.5 exposure increased airway inflammation, rest for 21 days reversed the airway inflammatory responses, which was associated with the induction of inhibitory memory alveolar macrophages (AMs). Similarly, polycyclic aromatic hydrocarbons (PAHs) in PM2.5 exposure and rest decreased pulmonary inflammation, accompanied by inhibitory memory AMs. Once AMs were depleted, pulmonary inflammation was aggravated. PAHs in PM2.5 promoted the secretion of IL-33 from airway epithelial cells via the aryl hydrocarbon receptor (AhR)/ARNT pathway. High-throughput mRNA sequencing revealed that PM2.5 exposure and rest drastically changed the mRNA profiles in AMs, which was largely rescued in IL-33-/- mice. Collectively, our results indicate that PM2.5 may mitigate pulmonary inflammation, which is mediated by inhibitory trained AMs via IL-33 production from epithelial cells through the AhR/ARNT pathway. We provide the rationale that PM2.5 plays complicated roles in respiratory disease.

CD146 deficiency promotes inflammatory type 2 responses in pulmonary cryptococcosis.

Cryptococcus neoformans (C. neoformans) is an important opportunistic fungal pathogen for pulmonary cryptococcosis. Previously, we demonstrated that CD146 mediated the adhesion of C. neoformans to the airway epithelium. CD146 is more than an adhesion molecule. In the present study, we aimed to explore the roles of CD146 in the inflammatory response in pulmonary cryptococcosis. CD146 was decreased in lung tissues from patients with pulmonary cryptococcosis. Similarly, C. neoformans reduced pulmonary CD146 expression in mice following intratracheal inoculation. To explore the pathological roles of CD146 reduction in pulmonary cryptococcosis, CD146 knockout (KO) mice were inoculated with C. neoformans via intratracheal instillation. CD146 deficiency aggravated C. neoformans infection, as evidenced by a shortened survival time and increased fungal burdens in the lung. Inflammatory type 2 cytokines (IL-4, IL-5, and TNF-α) and alternatively activated macrophages were increased in the pulmonary tissues of CD146 KO-infected mice. CD146 is expressed in immune cells (macrophages, etc.) and nonimmune cells, i.e., epithelial cells and endothelial cells. Bone marrow chimeric mice were established and infected with C. neoformans. CD146 deficiency in immune cells but not in nonimmune cells increased fungal burdens in the lung. Mechanistically, upon C. neoformans challenge, CD146 KO macrophages produced more neutrophil chemokine KC and inflammatory cytokine TNF-α. Meanwhile, CD146 KO macrophages decreased the fungicidity and production of reactive oxygen species. Collectively, C. neoformans infection decreased CD146 in pulmonary tissues, leading to inflammatory type 2 responses, while CD146 deficiency worsened pulmonary cryptococcosis.

A comparative study of IL-33 and its receptor ST2 in a C57BL/6 J mouse model of pulmonary Cryptococcus neoformans infection.

It has been reported that IL-33 receptor ST2 deficiency mitigates Cryptococcus neoformans (C. neoformans) pulmonary infection in BALB/c mice. IL-33 may modulate immune responses in ST2-dependent and ST2-independent manners. The host genetic background (i.e., BALB/c, C57BL/6 J) influences immune responses against C. neoformans. In the present study, we aimed to explore the roles of IL-33 and ST2 in pulmonary C. neoformans-infected mice on a C57BL/6 J genetic background. C. neoformans infection increased IL-33 expression in lung tissues. IL-33 deficiency but not ST2 deficiency significantly extended the survival time of C. neoformans-infected mice. In contrast, either IL-33 or ST2 deficiency reduced fungal burdens in lung, spleen and brain tissues from the mice following C. neoformans intratracheal inoculation. Similarly, inflammatory responses in the lung tissues were more pronounced in both the IL-33-/- and ST2-/- infected mice. However, mucus production was decreased in IL-33-/- infected mice alone, and the level of IL-5 in bronchoalveolar lavage fluid (BALF) was substantially decreased in the IL-33-/- infected mice but not ST2-/- infected mice. Moreover, IL-33 deficiency but not ST2 deficiency increased iNOS-positive macrophages. At the early stage of infection, the reduced pulmonary fungal burden in the IL-33-/- and ST2-/- mice was accompanied by increased neutrophil infiltration. Collectively, IL-33 regulated pulmonary C. neoformans infection in an ST2-dependent and ST2-independent manner in C57BL/6 J mice.

ResNet-based image inpainting method for enhancing the imaging speed of single molecule localization microscopy.

Single molecule localization microscopy (SMLM) is a mainstream method in the field of super-resolution fluorescence microscopy that can achieve a spatial resolution of 20∼30 nm through a simple optical system. SMLM usually requires thousands of raw images to reconstruct a super-resolution image, and thus suffers from a slow imaging speed. Recently, several methods based on image inpainting have been developed to enhance the imaging speed of SMLM. However, these image inpainting methods may also produce erroneous local features (or called image artifacts), for example, incorrectly joined or split filaments. In this study, we use the ResNet generator, a network with strong local feature extraction capability, to replace the popularly-used U-Net generator to minimize the image artifact problem in current image inpainting methods, and develop an image inpainting method called DI-STORM. We validate our method using both simulated and experimental data, and demonstrate that DI-STORM has the best acceleration capability and produces the least artifacts in the repaired images, as compared with VDSR (the simplest CNN-based image inpainting method in SMLM) and ANNA-PALM (the best GAN-based image inpainting method in SMLM). We believe that DI-STORM could facilitate the application of deep learning-based image inpainting methods for SMLM.

Selecting Multiple Node Statistics Jointly from Functional Connectivity Networks for Brain Disorders Identification.

Functional connectivity networks (FCN) analysis is instructive for the diagnosis of brain diseases, such as mild cognitive impairment (MCI) and major depressive disorder (MDD) at their early stages. As the critical step of FCN analysis, feature representation provides the basis for finding potential biomarkers of brain diseases. In previous studies, different node statistics (e.g. local efficiency and local clustering coefficients) are usually extracted from FCNs as features for the diagnosis/classification task, which can specifically locate disease-related regions on the node level, so as to help us understand the neurodevelopmental roots of brain disorders. However, each node statistic is proposed only considering a kind of specific network property, which has one-sidedness and limitations. As a result, it is incomplete to represent a node with only one statistic. To resolve this issue, we put forward a novel scheme to select multiple node statistics jointly from the estimated FCNs for automated classification, called multiple node statistics feature selection (MNSFS). Specifically, we first extract multiple statistics from the same nodes and assign each kind of statistic into a group. Then, sparse group least absolute shrinkage and selection operator (sgLASSO) is used to select groups (nodes) and statistics in the groups towards a better classification performance. Such a technique enables us to simultaneously locate the discriminative brain regions, as well as the specific statistics associated with these brain regions, making the classification results more interpretable. We conducted our scheme on two public databases for identifying subjects with MCI and MDD from normal controls. Experimental results show that the proposed scheme achieves superior classification accuracy and features interpreted on the benchmark datasets.

Accelerating multi-emitter localization in super-resolution localization microscopy with FPGA-GPU cooperative computation.

The real-time multi-emitter localization method is essential for advancing high-throughput super-resolution localization microscopy (HT-SRLM). In the past decade, the graphics processing unit (GPU) computation has been dominantly used to accelerate the execution speed of the multi-emitter localization method. However, if HT-SRLM is combined with a scientific complementary metal-oxide-semiconductor (sCMOS) camera working at full frame rate, real-time image processing is still difficult to achieve using this acceleration approach, thus resulting in a massive data storage challenge and even system crash. Here we take advantage of the cooperative acceleration power of field programming gate array (FPGA) computation and GPU computation, and propose a method called HCP-STORM to enable real-time multi-emitter localization. Using simulated images, we verified that HCP-STORM is capable of providing real-time image processing for raw images from a representative Hamamatsu Flash 4 V3 sCMOS camera working at full frame rate (that is, 2048×2048 pixels @ 10 ms exposure time). Using experimental images, we prove that HCP-STORM is 25 times faster than QC-STORM and 295 times faster than ThunderSTORM, with a small but acceptable degradation in image quality. This study shows the potential of FPGA-GPU cooperative computation in accelerating multi-emitter localization, and pushes a significant step toward the maturity of HT-SRLM technology.

A novel microRNA miR-1165-3p as a potential diagnostic biomarker for allergic asthma.

CONTEXT: A further examination of a novel miRNA,miR-1165-3p as a biomarker for asthma, which was previously implicated in helper T cells (Th2) in a murine asthma model. OBJECTIVE: To determine whether serum miR-1165-3p can serve as a potential diagnostic biomarker for allergic asthma. METHODS: Serum miR-1165-3p was quantified via quantitative real-time PCR (qRT-PCR) in asthmatic and control samples. Serum miR-1165-3p levels were compared between groups and the clinical diagnostic abilities of miR-1165-3p were evaluated. The analyses utilized included a student's t test, one-way ANOVA, and the generation of receiver operating characteristic (ROC) curves. RESULTS: Serum miRNA-1165-3p levels were significantly elevated in asthmatics when compared to the healthy controls. Furthermore, the sensitivity and specificity of serum miR-1165-3p were found to be 83% and 68.2%. Additionally, serum miR-1165-3p levels were also found to be significantly elevated in patients with allergic rhinitis (AR) or allergic bronchopulmonary aspergillosis (ABPA). CONCLUSIONS: This study showed that serum miR-1165-3p can potentially be utilized as a noninvasive biomarker that is able to aid in the diagnosis and characterization of allergic asthma.

Inhibition of H3K27me3 demethylases attenuates asthma by reversing the shift in airway smooth muscle phenotype.

BACKGROUND: The shift in airway smooth muscle cells (ASMCs) phenotype between proliferation and contraction during asthma has been reported recently, highlighting a role of ASMCs plasticity in the pathophysiology of asthma. As an event involved in epigenetic post-translational modification, histone H3 lysine27 (H3K27) demethylation has attracted significant attention with respect to the epigenetic changes in diverse cells; however, little is known about its contribution to the switching of ASMCs phenotype in asthma. OBJECTIVE: To investigate the role of trimethylated H3K27 (H3k27me3) demethylation in ASM remodelling as well as the underling mechanism. METHODS: Mice were exposed five times a week to house dust mite (HDM) extract for 5 weeks. Lung function was measured following the final HDM challenge. Airway inflammation and remodelling were then assessed in lungs of individual mice. Human ASMCs were purchased from Sciencell Research Laboratories. Proliferation, synthesis, migration and contraction of ASMCs were analysed, respectively. RESULTS: We observed demethylation at H3k27me3 sites in lungs harvested from mice exposed to HDM extract. Administration of a selective inhibitor of H3K27 demethylase (GSK-J4) could ameliorate the classical hallmarks of asthma, such as airway hyperresponsiveness, airway inflammation and remodelling. We established a proliferative as well as a contractive model of human ASMCs to explore the impacts of H3K27 demethylase inhibition on ASMCs phenotype. Our results indicated that GSK-J4 decreased ASMCs proliferation and migration elicited by PDGF through the Akt/JNK signalling; GSK-J4 also prevented the upregulation of contractile proteins in ASMCs induced by TGF-ß through the Smad3 pathway. CONCLUSIONS: Inhibition of H3K27me3 demethylation alleviated the development of asthmatic airway disease in vivo and modulated ASMCs phenotype in vitro. Collectively, our findings highlight a role of H3K27me3 demethylation in experimental asthma and ASMCs phenotype switch.

Robust multi-atlas label propagation by deep sparse representation.

Recently, multi-atlas patch-based label fusion has achieved many successes in medical imaging area. The basic assumption in the current state-of-the-art approaches is that the image patch at the target image point can be represented by a patch dictionary consisting of atlas patches from registered atlas images. Therefore, the label at the target image point can be determined by fusing labels of atlas image patches with similar anatomical structures. However, such assumption on image patch representation does not always hold in label fusion since (1) the image content within the patch may be corrupted due to noise and artifact; and (2) the distribution of morphometric patterns among atlas patches might be unbalanced such that the majority patterns can dominate label fusion result over other minority patterns. The violation of the above basic assumptions could significantly undermine the label fusion accuracy. To overcome these issues, we first consider forming label-specific group for the atlas patches with the same label. Then, we alter the conventional flat and shallow dictionary to a deep multi-layer structure, where the top layer (label-specific dictionaries) consists of groups of representative atlas patches and the subsequent layers (residual dictionaries) hierarchically encode the patchwise residual information in different scales. Thus, the label fusion follows the representation consensus across representative dictionaries. However, the representation of target patch in each group is iteratively optimized by using the representative atlas patches in each label-specific dictionary exclusively to match the principal patterns and also using all residual patterns across groups collaboratively to overcome the issue that some groups might be absent of certain variation patterns presented in the target image patch. Promising segmentation results have been achieved in labeling hippocampus on ADNI dataset, as well as basal ganglia and brainstem structures, compared to other counterpart label fusion methods.

LincR-PPP2R5C regulates IL-1ß ubiquitination in macrophages and promotes airway inflammation and emphysema in a murine model of COPD.

Chronic obstructive pulmonary disease (COPD) is a common disease with high global morbidity and mortality. Macrophages release IL-1ß and orchestrate airway inflammation in COPD. Previously, we explored the role of a new lncRNA, LincR-PPP2R5C, in regulating Th2 cells in asthma. Here, we established a murine model of COPD and explored the roles and mechanisms by which LincR-PPP2R5C regulates IL-1ß in macrophages. LincR-PPP2R5C was highly expressed in pulmonary macrophages from COPD-like mice. LincR-PPP2R5C deficiency ameliorated emphysema and pulmonary inflammation, as characterized by reduced IL-1ß in macrophages. Unexpectedly, in both lung tissues and macrophages, LincR-PPP2R5C deficiency decreased the expression of the IL-1ß protein but not the IL-1ß mRNA. Furthermore, we found that LincR-PPP2R5C deficiency increased the level of ubiquitinated IL-1ß in macrophages, which was mediated by PP2A activity. Targeting PP2A with FTY720 decreased IL-1ß and improved COPD. In conclusion, LincR-PPP2R5C regulates IL-1ß ubiquitination by affecting PP2A activity in macrophages, contributing to the airway inflammation and emphysema in a murine model of COPD. PP2A and IL-1ß ubiquitination in macrophages might be new therapeutic avenues for COPD therapy.

Plasma sCD146 is a potential biomarker for acute exacerbation of chronic obstructive pulmonary disease.

This study examined the levels of soluble CD146 (sCD146) in plasma samples from patients with chronic obstructive pulmonary disease (COPD) and assessed the relationship between sCD146 and the severity of COPD. A total of 97 COPD patients were recruited from 20 medical centers in Jiangsu, China, including 13 stable subjects and 84 exacerbated subjects. The plasma sCD146 level in exacerbated subjects (28.77 ± 10.80 ng/mL) was significantly lower than that in stable subjects (38.84 ± 15.00 ng/mL). In the high sCD146 group, the proportion of subjects with modified Medical Research Council (mMRC) scores of 0-1 was higher, the proportion of subjects with the Global Initiative for Chronic Obstructive Lung Disease (GOLD) stage 4 was lower, and the proportion of subjects with ≥1 hospitalizations in the past year was lower. The plasma sCD146 level was negatively correlated with the COPD Assessment Test (CAT) score (r = -0.2664, p = 0.0087). Logistic regression analysis showed that sCD146 was an independent risk factor for acute exacerbation of COPD (AECOPD). Receiver operating characteristic (ROC) analysis suggested that sCD146 combined with sex, age, pulmonary function, and acute exacerbations in the past year had clinical value for the accurate identification of AECOPD, with an area under the ROC curve (AUC) of 0.908 (95% CI: 0.810-1.000, p < 0.001). In addition, there was a significant negative correlation between plasma sCD146 and S100A9 (r = -0.3939, p < 0.001).

CD146 deficiency aggravates chronic obstructive pulmonary disease via the increased production of S100A9 and MMP-9 in macrophages.

Chronic obstructive pulmonary disease (COPD) is a leading cause of global death. As a molecule beyond adhesion, CD146 is involved in COPD pathogenesis. However, the mechanisms of CD146 in COPD remain largely elusive. We hypothesized that CD146 regulates the production of matrix metalloproteinase-9 (MMP-9) in macrophages and thereby contributes to COPD. Here, we constructed a murine model of COPD using lipopolysaccharide (LPS) and porcine pancreatic elastase (PPE). In COPD-like mice, LPS and PPE decreased the pulmonary expression of CD146. MMP-9 expression and bioactivity were increased in CD146 knockout COPD-like mice. In vitro, LPS decreased CD146 expression in macrophages. With or without LPS challenge, CD146-defective macrophages produced more MMP-9. Transcriptome analysis based on next-generation sequencing (NGS) revealed that S100A9 regulated MMP-9 production in CD146-defective macrophages. Targeting S100A9 with paquinimod decreased lung inflammation and alleviated alveolar destruction in COPD-like mice. Collectively, our study suggests that CD146 negatively regulates MMP-9 production in macrophages via the S100A9 pathway in COPD.

TOM5 regulates the mitochondrial membrane potential of alveolar epithelial cells in organizing pneumonia.

Deficiency of TOM5, a mitochondrial protein, causes organizing pneumonia (OP) in mice. The clinical significance and mechanisms of TOM5 in the pathogenesis of OP remain elusive. We demonstrated that TOM5 was significantly increased in the lung tissues of OP patients, which was positively correlated with the collagen deposition. In a bleomycin-induced murine model of chronic OP, increased TOM5 was in line with lung fibrosis. In vitro, TOM5 regulated the mitochondrial membrane potential in alveolar epithelial cells. TOM5 reduced the proportion of early apoptotic cells and promoted cell proliferation. Our study shed light on the roles of TOM5 in OP.

IFN-γ decreases PD-1 in T lymphocytes from convalescent COVID-19 patients via the AKT/GSK3ß signaling pathway.

Post-COVID-19 syndrome may be associated with the abnormal immune status. Compared with the unexposed age-matched elder group, PD-1 in the CD8+ T cells from recovered COVID-19 patients was significantly lower. IFN-γ in the plasma of COVID-19 convalescent patients was increased, which inhibited PD-1 expression in CD8+ T cells from COVID-19 convalescent patients. scRNA-seq bioinformatics analysis revealed that AKT/GSK3ß may regulate the INF-γ/PD-1 axis in CD8+ T cells from COVID-19 convalescent patients. In parallel, an IFN-γ neutralizing antibody reduced AKT and increased GSK3ß in PBMCs. An AKT agonist (SC79) significantly decreased p-GSK3ß. Moreover, AKT decreased PD-1 on CD8+ T cells, and GSK3ß increased PD-1 on CD8+ T cells according to flow cytometry analysis. Collectively, we demonstrated that recovered COVID-19 patients may develop long COVID. Increased IFN-γ in the plasma of recovered Wuhan COVID-19 patients contributed to PD-1 downregulation on CD8+ T cells by regulating the AKT/GSK3ß signaling pathway.

LincR-PPP2R5C Promotes Th2 Cell Differentiation Through PPP2R5C/PP2A by Forming an RNA-DNA Triplex in Allergic Asthma.

PURPOSE: The roles and mechanisms of long noncoding RNAs (lncRNAs) in T helper 2 (Th2) differentiation from allergic asthma are poorly understood. We aimed to explore a novel lncRNA, LincR-protein phosphatase 2 regulatory subunit B' gamma (PPP2R5C), in Th2 differentiation in a mouse model of asthma. METHODS: LincR-PPP2R5C from RNA-seq data of CD4+ T cells of asthma-like mice were validated and confirmed by quantitative reverse transcription polymerase chain reaction, northern blotting, nuclear and cytoplasmic separation, and fluorescence in situ hybridization (FISH). Lentiviruses encoding LincR-PPP2R5C or shRNA were used to overexpress or silence LincR-PPP2R5C in CD4+ T cells. The interactions between LincR-PPP2R5C and PPP2R5C were explored with western blotting, chromatin isolation by RNA purification assay, and fluorescence resonance energy transfer. An ovalbumin-induced acute asthma model in knockout (KO) mice (LincR-PPP2R5C KO, CD4 conditional LincR-PPP2R5C KO) was established to explore the roles of LincR-PPP2R5C in Th2 differentiation. RESULTS: LncR-PPP2R5C was significantly higher in CD4+ T cells from asthmatic mice ex vivo and Th2 cells in vitro. The lentivirus encoding LincR-PPP2R5C suppressed Th1 differentiation; in contrast, the short hairpin RNA (shRNA) lentivirus decreased LincR-PPP2R5C and Th2 differentiation. Mechanistically, LincR-PPP2R5C deficiency suppressed the phosphatase activity of the protein phosphatase 2A (PP2A) holocomplex, resulting in a decline in Th2 differentiation. The formation of an RNA-DNA triplex between LincR-PPP2R5C and the PPP2R5C promoter enhanced PPP2R5C expression and activated PP2A. LincR-PPP2R5C KO and CD4 conditional KO decreased Th2 differentiation, airway hyperresponsiveness and inflammatory responses. CONCLUSIONS: LincR-PPP2R5C regulated PPP2R5C expression and PP2A activity by forming an RNA-DNA triplex with the PPP2R5C promoter, leading to Th2 polarization in a mouse model of acute asthma. Our data presented the first definitive evidence of lncRNAs in the regulation of Th2 cells in asthma.

Deep-Learning-Based COVID-19 Diagnosis and Implementation in Embedded Edge-Computing Device.

The rapid spread of coronavirus disease 2019 (COVID-19) has posed enormous challenges to the global public health system. To deal with the COVID-19 pandemic crisis, the more accurate and convenient diagnosis of patients needs to be developed. This paper proposes a deep-learning-based COVID-19 detection method and evaluates its performance on embedded edge-computing devices. By adding an attention module and mixed loss into the original VGG19 model, the method can effectively reduce the parameters of the model and increase the classification accuracy. The improved model was first trained and tested on the PC X86 GPU platform using a large dataset (COVIDx CT-2A) and a medium dataset (integrated CT scan); the weight parameters of the model were reduced by around six times compared to the original model, but it still approximately achieved 98.80%and 97.84% accuracy, outperforming most existing methods. The trained model was subsequently transferred to embedded NVIDIA Jetson devices (TX2, Nano), where it achieved 97% accuracy at a 0.6-1 FPS inference speed using the NVIDIA TensorRT engine. The experimental results demonstrate that the proposed method is practicable and convenient; it can be used on a low-cost medical edge-computing terminal. The source code is available on GitHub for researchers.

Spatial-temporal data-augmentation-based functional brain network analysis for brain disorders identification.

Introduction: Due to the lack of devices and the difficulty of gathering patients, the small sample size is one of the most challenging problems in functional brain network (FBN) analysis. Previous studies have attempted to solve this problem of sample limitation through data augmentation methods, such as sample transformation and noise addition. However, these methods ignore the unique spatial-temporal information of functional magnetic resonance imaging (fMRI) data, which is essential for FBN analysis. Methods: To address this issue, we propose a spatial-temporal data-augmentation-based classification (STDAC) scheme that can fuse the spatial-temporal information, increase the samples, while improving the classification performance. Firstly, we propose a spatial augmentation module utilizing the spatial prior knowledge, which was ignored by previous augmentation methods. Secondly, we design a temporal augmentation module by random discontinuous sampling period, which can generate more samples than former approaches. Finally, a tensor fusion method is used to combine the features from the above two modules, which can make efficient use of spatial-temporal information of fMRI simultaneously. Besides, we apply our scheme to different types of classifiers to verify the generalization performance. To evaluate the effectiveness of our proposed scheme, we conduct extensive experiments on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset and REST-meta-MDD Project (MDD) dataset. Results: Experimental results show that the proposed scheme achieves superior classification accuracy (ADNI: 82.942%, MDD: 63.406%) and feature interpretation on the benchmark datasets. Discussion: The proposed STDAC scheme, utilizing both spatial and temporal information, can generate more diverse samples than former augmentation methods for brain disorder classification and analysis.

Deep learning using a residual deconvolutional network enables real-time high-density single-molecule localization microscopy.

High-density localization based on deep learning is a very effective method to accelerate single molecule localization microscopy (SMLM). Compared with traditional high-density localization methods, deep learning-based methods enable a faster data processing speed and a higher localization accuracy. However, the reported high-density localization methods based on deep learning are still not fast enough to enable real time data processing for large batches of raw images, which is probably due to the heavy computational burden and computation complexity in the U-shape architecture used in these models. Here we propose a high-density localization method called FID-STORM, which is based on an improved residual deconvolutional network for the real-time processing of raw images. In FID-STORM, we use a residual network to extract the features directly from low-resolution raw images rather than the U-shape network from interpolated images. We also use a model fusion from TensorRT to further accelerate the inference of the model. In addition, we process the sum of the localization images directly on GPU to obtain an additional speed gain. Using simulated and experimental data, we verified that the FID-STORM method achieves a processing speed of 7.31 ms/frame at 256 × 256 pixels @ Nvidia RTX 2080 Ti graphic card, which is shorter than the typical exposure time of 10∼30 ms, thus enabling real-time data processing in high-density SMLM. Moreover, compared with a popular interpolated image-based method called Deep-STORM, FID-STORM enables a speed gain of ∼26 times, without loss of reconstruction accuracy. We also provided an ImageJ plugin for our new method.

FOXA2 plays a critical role in hepatocellular carcinoma progression and lenvatinib-associated drug resistance.

Hepatic forkhead box protein A2 (FOXA2) is a crucial transcription factor for liver development and metabolic homeostasis. However, its role in hepatocellular carcinoma (HCC) progression and lenvatinib-related drug resistance remains unknown. In this study, the level of FOXA2 expression was found to be lower in HCC tissues than in paired adjacent tumor tissues. A low level of FOXA2 expression was associated with aggressive tumor characteristics (vascular invasion and poor differentiation). A low level of FOXA2 expression was found to be an independent risk factor for tumor recurrence (hazard ratio (HR): 1.899, P < 0.001) and long-term survival (HR: 2.011, P = 0.003) in HCC patients after hepatectomy. In xenograft animal models, FOXA2 overexpression significantly inhibited tumor growth. Moreover, FOXA2 overexpression was found to enhance the inhibitory effect of lenvatinib on HCC cells by upregulating the adenosine monophosphate-activated protein kinase-mechanistic target of rapamycin (AMPK-mTOR) pathway. Conversely, inhibition of adenosine monophosphate-activated protein kinase (AMPK) or stimulation of mechanistic target of rapamycin (mTOR) attenuated the sensitization of cells overexpressing FOXA2 to lenvatinib. Similarly, FOXA2 overexpression augmented the antitumor effect of lenvatinib in animal models with xenograft tumors. FOXA2 overexpression increased autophagy in HCC cells treated with lenvatinib. Lenvatinib treatment activated the platelet-derived growth factor receptor-extracellular regulated protein kinase (PDGFR-ERK) pathway in HCC. FOXA2 overexpression further downregulated the PDGFR-ERK pathway through the activation of the AMPK-mTOR axis. In conclusion, FOXA2 was identified as an independent risk factor for HCC after hepatectomy. FOXA2 was found to be closely associated with the biological progression of HCC. By modulating the AMPK-mTOR-autophagy signaling pathway, FOX2 significantly augmented antitumor effect of lenvatinib in HCC.