Robust Sensory Information Reconstruction and Classification With Augmented Spikes.

Sensory information recognition is primarily processed through the ventral and dorsal visual pathways in the primate brain visual system, which exhibits layered feature representations bearing a strong resemblance to convolutional neural networks (CNNs), encompassing reconstruction and classification. However, existing studies often treat these pathways as distinct entities, focusing individually on pattern reconstruction or classification tasks, overlooking a key feature of biological neurons, the fundamental units for neural computation of visual sensory information. Addressing these limitations, we introduce a unified framework for sensory information recognition with augmented spikes. By integrating pattern reconstruction and classification within a single framework, our approach not only accurately reconstructs multimodal sensory information but also provides precise classification through definitive labeling. Experimental evaluations conducted on various datasets including video scenes, static images, dynamic auditory scenes, and functional magnetic resonance imaging (fMRI) brain activities demonstrate that our framework delivers state-of-the-art pattern reconstruction quality and classification accuracy. The proposed framework enhances the biological realism of multimodal pattern recognition models, offering insights into how the primate brain visual system effectively accomplishes the reconstruction and classification tasks through the integration of ventral and dorsal pathways.

Association of blood pressure and outcomes differs upon cerebral perfusion post-thrombectomy in patients with acute ischemic stroke.

BACKGROUND: The relationship between post-endovascular thrombectomy (EVT) blood pressure (BP) and outcomes in patients with acute ischemic stroke (AIS) remains contentious. We aimed to explore whether this association differs with different cerebral perfusion statuses post-EVT. METHODS: In a multicenter observational study of patients with AIS with large vessel occlusion who underwent EVT, we enrolled those who accepted CT perfusion (CTP) imaging within 24 hours post-EVT. We recorded post-EVT systolic (SBP) and diastolic BP. Patients were stratified into favorable perfusion and unfavorable perfusion groups based on the hypoperfusion intensity ratio (HIR) on CTP. The primary outcome was good functional outcome (90-day modified Rankin Scale score of ≤3). Secondary outcomes included early neurological deterioration, infarct size growth, and symptomatic intracranial hemorrhage. RESULTS: Of the 415 patients studied (mean age 62 years, 75% male), 233 (56%) achieved good functional outcomes. Logistic regression showed that post-EVT HIR and 24-hour mean SBP were significantly associated with functional outcomes. Among the 326 (79%) patients with favorable perfusion, SBP <140 mmHg was associated with a higher percentage of good functional outcomes compared with SBP ≥140 mmHg (68% vs 52%; aOR 1.70 (95% CI 1.00 to 2.89), P=0.04). However, no significant difference was observed between SBP and functional outcomes in the unfavorable perfusion group. There was also no discernible difference between SBP and secondary outcomes across the different perfusion groups. CONCLUSIONS: In patients with favorable perfusion post-EVT, SBP <140 mmHg was associated with good functional outcomes, which underscores the need for further investigations with larger sample sizes or a more individualized BP management strategy. CLINICAL TRIAL REGISTRATION: ChiCTR1900022154.

The cost-effectiveness of including liquid biopsy into molecular profiling strategies for newly diagnosed advanced non-squamous non-small cell lung cancer in an Asian population.

OBJECTIVES: Liquid biopsy is complementary to tissue biopsy for lung cancer profiling, yet evidence of the cost-effectiveness is limited. This could retard implementation and reimbursement in clinical practice. The aim of this study is to estimate the cost-effectiveness of profiling strategies that include liquid biopsy and to identify the optimal profiling approach for newly diagnosed advanced non-squamous non-small cell lung cancer (NSCLC) in an Asian population using Singapore as an example. MATERIALS AND METHODS: A decision tree and partitioned-survival model was developed from the Singapore healthcare system's perspective to evaluate the cost-effectiveness of five molecular profiling strategies: either tissue or plasma next-generation sequencing (NGS) alone, a concurrent, and two sequential approaches. Model inputs were informed by local data or published literature. Sensitivity analyses and scenario analyses were undertaken to understand the robustness of the conclusions for decision making. The optimal strategy at different willingness-to-pay (WTP) thresholds was presented by cost-effectiveness acceptability frontier and the expected loss curve. RESULTS: The sequential tissue-plasma NGS approach revealed an additional 0.0981 quality adjusted life years (QALYs) for an extra cost of S$3,074 over a 20-year time horizon compared to tissue NGS alone, resulting in an incremental cost-effectiveness ratio (ICER) of S$31,318/QALY and an incremental net monetary benefit of S$1,343 per patient. The findings were sensitive to the costs of pembrolizumab and osimertinib and the probabilities of re-biopsy after tissue NGS. Sequential plasma-tissue NGS and plasma NGS alone were more costly and less effective than alternatives. CONCLUSION: The sequential tissue-plasma NGS approach generated the highest net monetary benefit and was the optimal testing strategy when WTP was S$45,000/QALY. It retained superiority but understandably with a higher ICER when expensive, non-first line treatments were included. Overall, its routine clinical practice should be proactively considered for newly diagnosed advanced non-squamous NSCLC in an Asian population.

Effect of exopolysaccharides yield and addition concentration of Lactobacillus helveticus on the processing characteristics of fermented milk and its mechanism.

Exopolysaccharides (EPS) yield and added concentration of lactic acid bacteria can greatly affect the processing characteristics of fermented milk. In order to investigate the effects and mechanisms of EPS yield and added concentration on fermented milk, researchers extracted EPS from 50 strains of Lactobacillus helvedicus (L. helvedicus) and selected the two strains with the largest difference in EPS yield (L. helvedicus LH18 and L. helvetigus LH33) for subsequent experiments. The physicochemical properties of EPS-LH18 and EPS-LH33 were analyzed. The gel characteristics and protein conformation of fermented milk were studied by means of texture analyzer, rheometer, scanning electron microscopy, nuclear magnetic resonance machine, fluorescence spectrophotometer and circular dichroism. The results indicate that the monosaccharide compositions of EPS-LH18 and EPS-LH33 are the same and have good thermal stability. The texture and rheological properties of L. helveticus LH18 fermented milk are significantly superior to other fermented milk. The reason is that L. helveticus LH18 EPS has the highest yield, which leads to a denser gel structure, lower surface hydrophobicity and free sulfhydryl content of its fermented milk. According to circular dichroism analysis, ß- sheet and random coil are the internal factors leading to the difference in fermented milk gel. In addition, the fermented milk improved even more favorably as the concentration of the two EPS additions increased. As described above, L. helveticus LH18 has the potential to be an excellent yogurt starter, and both of the above EPS can be used as probiotic stabilizer alternatives for fermented dairy products.

Exopolysaccharides of lactic acid bacteria: Structure, biological activity, structure-activity relationship, and application in the food industry: A review.

Over the past few decades, researchers have discovered that probiotics play an important role in our daily lives. With the further deepening of research, more and more evidence show that bacterial metabolites have an important role in food and human health, which opens up a new direction for the research of lactic acid bacteria (LAB) in the food and pharmaceutical industry. Many LAB have been widely studied because of the ability of exopolysaccharides (EPS). Lactic acid bacteria exopolysaccharides (LAB EPS) not only have great potential in the treatment of human diseases but also can become natural ingredients in the food industry to provide special qualitative structure and flavor. This paper has organized and summarized the biosynthesis, strain selection, production process parameters, structure, and biological activity of LAB EPS, filling in the monotony and incompleteness of previous articles' descriptions of LAB EPS. Therefore, this paper focuses on the general biosynthetic pathway, structural characterization, structure-activity relationship, biological activity of LAB EPS, and their application in the food industry, which will help to deepen people's understanding of LAB EPS and develop new active drugs from LAB EPS. Although the research results are relatively affluent, the low yield, complex structure, and few clinical trials of EPS are still the reasons that hinder its development. Therefore, future knowledge expansion should focus on the regulation of structure, physicochemical properties, function, higher production of EPS, and clinical trial applications, which can further increase the commercial significance and value of EPS. Furthermore, better understanding the structure-function relationship of EPS in food remains a challenge to date.

LYSTO: The Lymphocyte Assessment Hackathon and Benchmark Dataset.

We introduce LYSTO, the Lymphocyte Assessment Hackathon, which was held in conjunction with the MICCAI 2019 Conference in Shenzhen (China). The competition required participants to automatically assess the number of lymphocytes, in particular T-cells, in images of colon, breast, and prostate cancer stained with CD3 and CD8 immunohistochemistry. Differently from other challenges setup in medical image analysis, LYSTO participants were solely given a few hours to address this problem. In this paper, we describe the goal and the multi-phase organization of the hackathon; we describe the proposed methods and the on-site results. Additionally, we present post-competition results where we show how the presented methods perform on an independent set of lung cancer slides, which was not part of the initial competition, as well as a comparison on lymphocyte assessment between presented methods and a panel of pathologists. We show that some of the participants were capable to achieve pathologist-level performance at lymphocyte assessment. After the hackathon, LYSTO was left as a lightweight plug-and-play benchmark dataset on grand-challenge website, together with an automatic evaluation platform.

Milk fat globule membrane regulates the physicochemical properties and surface composition of infant formula powders by improving the stability of the emulsion.

The milk fat globules in infant formula (IF) are encapsulated by a component known as milk fat globule membrane (MFGM). However, it is currently unclear whether the improved emulsion stability of MFGM can have a profound effect on the finished IF. Therefore, this study investigated the effects of MFGM on the particle size, stability, rheology, and microstructure of emulsions prepared by dairy ingredients via wet mixing. Further, IF were processed using such emulsions, the physicochemical properties, surface composition of the powders were examined. The results showed that MFGM reduced the particle size of the emulsion, increased the viscosity, and improved the microstructure of the MFGM. Furthermore, MFGM reduced the moisture content of the powder, increased the glass transition temperature, and reduced the presence of surface fat. In conclusion, the addition of MFGM enhance the finished powder stability by improving the emulsion stability prepared during IF manufacturing.

Optimal systolic and diastolic blood pressure threshold that associated with lower risk of white matter hyperintensity progression.

Background: The optimal control thresholds for systolic blood pressure (SBP) and diastolic blood pressure (DBP) in patients with white matter hyperintensity (WMH) are still unclear. Method: A longitudinal retrospective study of patients with brain magnetic resonance imaging (MRI) scans with intervals of more than 3 years was conducted. Blood pressure records during hospitalization and from outpatient visits between baseline and the last MRI scan were collected. The outcome was the change in total WMH from baseline to the final visit. Results: Among the 965 patients with MRI scans, 457 patients with detailed longitudinal blood pressure records were ultimately included and classified into the WMH absent group (n = 121), mild WMH group (n = 126), and moderate to severe WMH group (n = 210). Both baseline and longitudinal mean SBP, DBP, and SBP SD were significantly associated with WMH severity (p < 0.05). An average SBP of 130-140 mmHg [vs. <130 mmHg, aOR, 1.80, (95% CI, 1.05-3.07), p = 0.03] was associated with a higher risk of WMH progression. DBP ≥ 90 mmHg [vs. <80 mmHg, OR, 1.81, (95% CI, 0.88-3.74), p = 0.02, aOR, 1.54, (95% CI, 0.66-3.53), p = 0.32] was associated with a higher risk of WMH progression, but was not after adjusted for other covariates. Longitudinal BP variability was not significantly associated with WMH progression. Conclusion: Both SBP and DBP had a stronger relationship with the severity of WMH. A target mean SBP of <130 mmHg and mean DBP of <80 mmHg was associated with a lower risk of WMH progression.

Economic Evaluations of Imaging Biomarker-Driven Companion Diagnostics for Cancer: A Systematic Review.

INTRODUCTION: There is a boom in imaging biomarker-driven companion and complementary diagnostics (CDx) for cancer, which brings opportunity for personalized medicine. Whether adoption of these technologies is likely to be cost-effective is a relevant question, and studies on this topic are emerging. Despite the growing number of economic evaluations, no review of the methods used, quality of reporting, and potential risk of bias has been done. We report a systematic review to identify, summarize, and critique the cost-effectiveness evidence for the use of biomarker-driven and imaging-based CDx to inform cancer treatments. METHODS: The Preferred Reporting Items of Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed. Systematic literature searches until 30 December 2022 were performed in PubMed, Web of Science, Medline, Embase, and Scopus for economic evaluations of imaging biomarker-based CDx for cancer. The inclusion and exclusion of studies were determined by pre-specified eligibility criteria informed by the 'Patient, Intervention, Comparison, Outcome' (PICO) framework. The Consolidated Health Economic Evaluation Reporting Standards (CHEERS) was used to assess the quality of reporting, and the Bias in Economic Evaluation (ECOBIAS) was used to examine the potential risk of bias of included studies. RESULTS: A total of 12 papers were included, with eight model-based and four trial-based studies. Implementing biomarker-driven, imaging-based CDx was reported to be cost-effective, cost saving, or dominant (cost saving and more effective) in ten papers. Inconsistent methods were found in the studies, and the quality of reporting was lacking against the CHEERS reporting guideline. Several potential sources of 'risk of bias' were identified. These should be acknowledged and carefully considered by researchers planning future health economic evaluations. CONCLUSION: Despite favorable results towards the implementation of imaging biomarker-based CDx for cancer, there is room for improvement regarding the quantity and quality of economic evaluations, and that is expected as the awareness of current study limitations increases and more clinical data become available in the future.

Study on the Characteristics of Early Cytokine Storm Response to Cardiac Surgery.

Cardiac surgery can provoke an acute cytokine storm that may contribute to the development of postoperative multiple organ dysfunction syndrome. We prospectively observed patients undergoing cardiac surgery and divided them into two groups: the severe group and the mild group. Healthy individuals were enrolled acting as the control group for comparison. Plasma samples and clinical data were recorded at the initiation of cardiac-pulmonary bypass (CPB) and 3, 6, 12, 24, and 48 h after initiation of CPB. Cytokine levels were detected using the Luminex® technique. Thirty-nine adults were enrolled in this study (14 in the severe group, 15 in the mild group, and 10 in the control group). Cytokine concentrations were significantly higher in the severe group. Principal component analysis was used to establish a cytokine storm intensity curve, which represented the overall trend of 10 cytokines. The peak concentrations of interleukin (IL)-6, IL-10, and IL-16 were 425.1, 198.5, and 623.0 pg/mL, which were more than 1,200, 1,800, and 240 times the normal level, respectively. The maximum cytokine storm intensity predated the maximum Vasoactive-Inotropic Score (VIS) and Sequential Organ Failure Assessment (SOFA) score in the severe group. Cytokine storm response to cardiac surgery occurred early and was associated with disease severity. Interventions to cytokine storm should be initiated early as guided by cytokine storm biomarkers such as IL-6, IL-10, and IL-16 in severe patients undergoing cardiac surgery. Clinical Trial Registration: ChiCTR1900021351.

A review of the signaling pathways of aerobic and anaerobic exercise on atherosclerosis.

Atherosclerosis (AS), a chronic inflammatory vascular disease with lipid metabolism abnormalities, is one of the major pathological bases of coronary heart disease. As people's lifestyles and diets change, the incidence of AS increases yearly. Physical activity and exercise training have recently been identified as effective strategies for lowering cardiovascular disease (CVD) risk. However, the best exercise mode to ameliorate the risk factors related to AS is not clear. The effect of exercise on AS is affected by the type of exercise, intensity, and duration. In particular, aerobic and anaerobic exercise are the two most widely discussed types of exercise. During exercise, the cardiovascular system undergoes physiological changes via various signaling pathways. The review aims to summarize signaling pathways related to AS in two different exercise types and provide new ideas for the prevention and treatment of AS in clinical practice.

An update on the role of TRIM/NLRP3 signaling pathway in atherosclerosis.

Atherosclerosis (AS) is a chronic inflammatory disease of large and medium arteries that includes lipid metabolism disorder and recruitment of immune cells to the artery wall. An increasing number of studies have confirmed that inflammasome over-activation is associated with the onset and progression of atherosclerosis. The NLRP3 inflammasome, in particular, has been proven to increase the incidence rate of cardiovascular diseases (CVD) by promoting pro-inflammatory cytokine release and reducing plaque stability. The strict control of inflammasome and prevention of excessive inflammatory reactions have been the research focus of inflammatory diseases. Tripartite motif (TRIM) is a protein family with a conservative structure and rapid evolution. Several studies have demonstrated the TRIM family's regulatory role in mediating inflammation. This review aims to clarify the relationship between TRIMs and NLRP3 inflammasome and provide insights for future research and treatment discovery.

Elucidating gut microbiota and metabolite patterns shaped by goat milk-based infant formula feeding in mice colonized by healthy infant feces.

The gut microbiota plays an evolutionarily conserved role in host metabolism, which is influenced by diet. Here, we investigated differences in shaping the gut microbiota and regulating metabolism in cow milk-based infant formula, goat milk-based infant formula, and mix milk-based infant formula compared with pasteurized human milk. 16S rRNA results showed that goat milk-based infant formula selectively increased the relative abundance of Blautia, Roseburia, Alistites and Muribaculum in the gut compared to other infant formulas. Metabolomics identification indicated that goat milk-based infant formula mainly emphasized bile acid biosynthesis, arachidonic acid metabolism and steroid biosynthesis metabolic pathways. Metabolites associated with these metabolic pathways were positively associated with increased microorganisms in goat milk-based infant formula, particularly Alistipes. Furthermore, we found a deficiency of Akkermansia abundance in three infant formula-fed compared to pasteurizedhuman milk-fed. This study presents new insights into the improvement and application of goat milk-based infant formulas in terms of intestinal microecology.

Effects of high-pressure homogenization on physicochemical and functional properties of enzymatic hydrolyzed soybean protein concentrate.

This paper investigates the effect on the physicochemical and functional properties of soybean protein concentrate (SPC) by using Alcalase protease and high-pressure homogenization (HPH) (0, 20, 40, 60, 80, and 100 MPa) for the combined modification. The results showed that the degree of hydrolysis of SPC was 4.1% and the antigen protein was degraded after Alcalase hydrolysis, when the homogenization pressure (HP) was 6 0Mpa, the particle size of the SPC was the smallest, the zate potential absolute value up to 33.45 mV, the secondary structure has the lowest ß-sheet content, the highest random coil content, and the highest surface hydrophobicity (H0), the size of protein fragments on the microstructure surface is the smallest, the lowest denaturation temperature (T d ) and enthalpy (△H) are 72.59°C and 1.35 J/g, the highest solubility is 80.54%, and the highest water and oil holding capacities are 7.73 g/g and 6.51 g/g, respectively. The best emulsifying activity and emulsifying stability were 43.46 m2/g and 190.35 min, the most even distribution of emulsion droplets. This indicates that the HPH treatment destroys the structure of enzymatic hydrolyzed SPC, changes its physicochemical properties, and improves its functional properties. In this study, SPC was modified by HPH and enzyme combined treatment, in order to improve the functionality and application range of SPC, and provide a theoretical basis for its high-value utilization in the food field.

Diversity investigation by application of DNA barcoding: A case study of lepidopteran insects in Xinjiang wild fruit forests, China.

To investigate the species diversity of lepidopteran insects in Xinjiang wild fruit forests, establish insect community monitoring systems, and determine the local species pool, we test the applicability of DNA barcoding based on cytochrome c oxidase subunit I (COI) gene for accurate and rapid identification of insect species. From 2017 to 2019, a total of 212 samples with ambiguous morphological identification were selected for DNA barcoding analysis. Five sequence-based methods for species delimitation (ABGD, BINs, GMYC, jMOTU, and bPTP) were conducted for comparison to traditional morphology-based identification. In total, 2,422 samples were recorded, representing 143 species of 110 genera in 17 families in Lepidoptera. The diversity analysis showed that the richness indices for Noctuidae was the highest (54 species), and for Pterophoridae, Cossidae, Limacodidae, Lasiocampidae, Pieridae, and Lycaenidae were the lowest (all with 1 species). The Shannon-Wiener species diversity index (H') and Pielou's evenness (J') of lepidopteran insects first increased and then decreased across these 3 years, while the Simpson diversity index showed a trend of subtracted then added. For molecular-based identification, 67 lepidopteran species within 61 genera in 14 families were identified through DNA barcoding. Neighbor-joining (NJ) analysis showed that conspecific individuals were clustered together and formed monophyletic groups with a high support value, except for Lacanobia contigua (Denis & Schiffermüller, 1775) (Noctuidae: Hadeninae). Sixty-seven morphospecies were classified into various numbers of MOTUs based on ABGD, BINs, GMYC, jMOTU, and bPTP (70, 96, 2, 71, and 71, respectively). In Xinjiang wild fruit forests, the family with the largest number of species is Noctuidae, followed by Geometridae, Crambidae, and the remaining families. The highest Shannon diversity index is observed for the family Noctuidae. Our results indicate that the distance-based methods (ABGD and jMOTU) and character-based method (bPTP) outperform GMYC. BINs is inclined to overestimate species diversity compared to other methods.

The price of political polarization: Evidence from municipal issuers during the coronavirus pandemic.

This study estimates the financial costs imposed by political polarization among citizens on U.S. local governments during the COVID-19 pandemic. We measure local political polarization by using citizens' voting results in the presidential elections. We find local political polarization leads to higher offering yield of the bonds issued by the U.S. municipalities. The impact on borrowing costs is exaggerated by the number of pandemic cases in the local area, suggesting political polarization hinders the making and enforcement of government measures for the pandemic. This study highlights the mechanisms through which financial markets and local political ideology jointly affect social welfare.

Investigating the impact of SARS-CoV-2 infection on basic semen parameters and in vitro fertilization/intracytoplasmic sperm injection outcomes: a retrospective cohort study.

BACKGROUND: This study aimed to evaluate the influences of SARS-CoV-2 infection on semen parameters and investigate the impact of the infection on in vitro fertilization (IVF) outcomes. METHODS: This retrospective study enrolled couples undergoing IVF cycles between May 2020 and February 2021 at Tongji Hospital, Wuhan. Baseline characteristics were matched using propensity score matching. Participants were categorized into an unexposed group (SARS-COV-2 negative) and exposed group (SARS-COV-2 positive) based on a history of SARS-CoV-2 infection, and the populations were 148 and 50 after matching, respectively. IVF data were compared between the matched cohorts. Moreover, semen parameters were compared before and after infection among the infected males. The main measures were semen parameters and IVF outcomes, including laboratory and clinical outcomes. RESULTS: Generally, the concentration and motility of sperm did not significantly differ before and after infection. Infected males seemed to have fewer sperm with normal morphology, while all values were above the limits. Notably, the blastocyst formation rate and available blastocyst rate in the exposed group were lower than those in the control group, despite similar mature oocytes rates, normal fertilization rates, cleavage rates, and high-quality embryo rates. Moreover, no significant differences were exhibited between the matched cohorts regarding the implantation rate, biochemical pregnancy rate, clinical pregnancy rate, or early miscarriage rate. CONCLUSIONS: The results of this retrospective cohort study suggested that the semen quality and the chance of pregnancy in terms of IVF outcomes were comparable between the males with a history of SARS-CoV-2 infection and controls, although a decreased blastocyst formation rate and available blastocyst rate was observed in the exposed group, which needs to be reinforced by a multicenter long-term investigation with a larger sample size.

STSRNet: Self-Texture Transfer Super-Resolution and Refocusing Network.

Biomedical microscopy images with high-resolution (HR) and axial information can help analysis and diagnosis. However, obtaining such images usually takes more time and economic costs, which makes it impractical in most scenarios. In this paper, we first propose a novel Self-texture Transfer Super-resolution and Refocusing Network (STSRNet) to reconstruct HR multi-focal plane (MFP) images from a single 2D low-resolution (LR) wide field image without relying on scanning or any special devices. The proposed STSRNet consists of three parts: the backbone module for extracting features, the self-texture transfer module for transferring and fusing features, and the flexible reconstruction module for SR and refocusing. Specifically, the self-texture transfer module is designed for images with self-similarity such as cytological images and it searches for similar textures within the image and transfers to help MFP reconstruction. As for reconstruction module, it is composed of multiple pluggable components, each of which is responsible for a specific focal plane, so as to performs SR and refocusing all focal planes at one time to reduce computation. We conduct extensive experiments on cytological images and the experiments show that MFP images reconstructed by STSRNet have richer details in the axial and horizontal directions than input images. At the same time, the reconstructed MFP images also perform better than single 2D wide field images on high-level tasks. The proposed method provides relatively high-quality MFP images when real MFP images cannot be obtained, which greatly expands the application potential of LR wide-field images. To further promote the development of this field, we released our cytology dataset named RSDC for more researchers to use.

Dynamics of Bacterial Community and Fermentation Quality in Leymus chinensis Silage Treated With Lactic Acid Bacteria and/or Water.

This study aimed to reveal the bacterial community and fermentation quality of Leymus chinensis silage during the fermentation process. L. chinensis was harvested at the heading stage, and ensiled with lactic acid bacteria (LAB, L), water (W), or a combination of both (LW) in vacuum-sealed plastic bags. As a control silage, untreated L. chinensis silage was also assessed. The samples were taken at 0, 5, 15, 35, and 60 days after ensiling. The bacterial community structure was assessed by plate cultivation and Illumina sequencing, and the fermentation parameters were also analyzed. Fresh L. chinensis contained low moisture (509 g/kg) and LAB (3.64 log colony-forming units/g fresh weight). Control silage displayed higher pH and lower lactic acid (LA) than other treatments during ensilage (p < 0.05); moreover, LW-treatment had lower pH from 5 to 35 days and greater LA at 5 days than L- and W-treatments (p < 0.05). During the fermentation process, Lactobacillus in L- and LW-treatments was the most dominant bacterial genus (>97%), had higher abundance than that in control silage and W-treatment (p < 0.05), and correlated negatively with other main genera and pH, and positively with LA and acetic acid (p < 0.05). Moreover, Lactobacillus had considerable abundance in W-treatment from 5 to 15 days (81.38-85.86%). Enterobacteriaceae had the most abundance among bacteria in control silage during ensiling (49.31-69.34%), and in W-treatment from 35 to 60 days (47.49-54.15%). The L-, W-, and LW-treatments displayed the aggregated bacterial community at 5 and 15 days, with W-treatment diverging from L- and LW-treatments at 35 and 60 days. Overall, the low moisture and/or insufficient LAB in fresh L. chinensis led to Enterobacteriaceae dominating bacterial community and contributing to the high pH and low LA in control silage during the fermentation process. Applying L, W, or LW contributed to Lactobacillus succession, LA production, and pH reduction during early stage of fermentation; moreover, treating with L and LW displayed more efficiency. Lactobacillus dominated the entire ensilage process in L- and LW-treatments and the early stage of fermentation in W-treatment, and contributed to the satisfactory fermentation quality of L. chinensis silage. The L- and LW-treatments displayed a similar pattern of bacterial succession during ensiling.

Robust whole slide image analysis for cervical cancer screening using deep learning.

Computer-assisted diagnosis is key for scaling up cervical cancer screening. However, current recognition algorithms perform poorly on whole slide image (WSI) analysis, fail to generalize for diverse staining and imaging, and show sub-optimal clinical-level verification. Here, we develop a progressive lesion cell recognition method combining low- and high-resolution WSIs to recommend lesion cells and a recurrent neural network-based WSI classification model to evaluate the lesion degree of WSIs. We train and validate our WSI analysis system on 3,545 patient-wise WSIs with 79,911 annotations from multiple hospitals and several imaging instruments. On multi-center independent test sets of 1,170 patient-wise WSIs, we achieve 93.5% Specificity and 95.1% Sensitivity for classifying slides, comparing favourably to the average performance of three independent cytopathologists, and obtain 88.5% true positive rate for highlighting the top 10 lesion cells on 447 positive slides. After deployment, our system recognizes a one giga-pixel WSI in about 1.5 min.