Analysis of CT and MR imaging features of the brain in patients with hydrogen sulfide poisoning based on clinical symptom grading.

OBJECTIVE: To retrospectively analyze CT and MR imaging features of the brain in patients with hydrogen sulfide poisoning based on clinical symptom grading and to investigate their correlations with clinical symptoms and patients' prognosis. METHODS: A retrospective analysis was performed of CT and MR imaging data of the brain in 40 patients with hydrogen sulfide poisoning in our hospital. There were four main imaging manifestations. Patients were clinically graded according to the central nervous system symptom scores of the Poisoning Severity Score (PSS) and staged according to the gas inhalation time segment. Based on clinical symptom grading, the frequencies and proportions of four imaging signs that occurred in each group were counted, their development trends were analyzed, and the correlations of imaging features with clinical grading and prognosis were calculated. RESULTS: Forty patients were divided into minor, moderate and severe clinical grades and classified into four stages. In patients with minor and moderate clinical grading, only one patient suffered from generalized brain edema at stage 1, with a good prognosis. Patients with severe clinical grade showed the highest probability of presenting with the four imaging signs. The imaging signs were correlated with the severe clinical grade and a poor prognosis (P = 0.000, R = 0.828; P = 0.000, R = 0.858). CONCLUSION: In patients with the severe clinical grade, generalized brain edema and symmetrical hypodensity/abnormal signals in the bilateral basal ganglia and around the lateral ventricles were the main findings and were shown to persist. The presence of imaging signs can assist in the clinically effective evaluation of clinical symptom grade.

IRCM-Caps: An X-ray image detection method for COVID-19.

OBJECTIVE: COVID-19 is ravaging the world, but traditional reverse transcription-polymerase reaction (RT-PCR) tests are time-consuming and have a high false-negative rate and lack of medical equipment. Therefore, lung imaging screening methods are proposed to diagnose COVID-19 due to its fast test speed. Currently, the commonly used convolutional neural network (CNN) model requires a large number of datasets, and the accuracy of the basic capsule network for multiple classification is limital. For this reason, this paper proposes a novel model based on CNN and CapsNet. METHODS: The proposed model integrates CNN and CapsNet. And attention mechanism module and multi-branch lightweight module are applied to enhance performance. Use the contrast adaptive histogram equalization (CLAHE) algorithm to preprocess the image to enhance image contrast. The preprocessed images are input into the network for training, and ReLU was used as the activation function to adjust the parameters to achieve the optimal. RESULT: The test dataset includes 1200 X-ray images (400 COVID-19, 400 viral pneumonia, and 400 normal), and we replace CNN of VGG16, InceptionV3, Xception, Inception-Resnet-v2, ResNet50, DenseNet121, and MoblieNetV2 and integrate with CapsNet. Compared with CapsNet, this network improves 6.96%, 7.83%, 9.37%, 10.47%, and 10.38% in accuracy, area under the curve (AUC), recall, and F1 scores, respectively. In the binary classification experiment, compared with CapsNet, the accuracy, AUC, accuracy, recall rate, and F1 score were increased by 5.33%, 5.34%, 2.88%, 8.00%, and 5.56%, respectively. CONCLUSION: The proposed embedded the advantages of traditional convolutional neural network and capsule network and has a good classification effect on small COVID-19 X-ray image dataset.

Structural elucidation of hemicelluloses from oil-tea camellia fruit shell.

Oil-tea camellia fruit shell (CFS) is a very abundant waste lignocellulosic resource. The current treatments of CFS, i.e. composting and burning, pose a severe threat on environment. Up to 50 % of the dry mass of CFS is composed of hemicelluloses. However, chemical structures of the hemicelluloses in CFS have not been extensively studied, which limits their high-value utilization. In this study, different types of hemicelluloses were isolated from CFS through alkali fractionation with the assistance of Ba(OH)2 and H3BO3. Xylan, galacto-glucomannan and xyloglucan were found to be the major hemicelluloses in CFS. Through methylation, HSQC and HMBC analyses, we have found that the xylan in CFS is composed of →4)-ß-D-Xylp-(1→ and →3,4)-ß-D-Xylp-(1→ linked by (1→4)-ß glycosidic bond as the main chain; the side chains are α-L-Fucp-(1→, →5)-α-L-Araf-(1→, ß-D-Xylp-(1→, α-L-Rhap-(1→ and 4-O-Me-α-D-GlcpA-(1→, connected to the main chain through (1→3) glycosidic bond. The main chain of galacto-glucomannan in CFS consists of →6)-ß-D-Glcp-(1→, →4)-ß-D-Glcp-(1→, →4,6)-ß-D-Glcp-(1→ and →4)-ß-D-Manp-(1→; the side chains are ß-D-Glcp-(1→, →2)-ß-D-Galp-(1→, ß-D-Manp-(1→ and →6)-ß-D-Galp-(1→ connected to the main chain through (1→6) glycosidic bonds. Moreover, galactose residues are connected by α-L-Fucp-(1→. The main chain of xyloglucan is composed of →4)-ß-D-Glcp-(1→, →4,6)-ß-D-Glcp-(1→ and →6)-ß-D-Glcp-(1→; the side groups, i.e. ß-D-Xylp-(1→ and →4)-ß-D-Xylp-(1→, are connected to the main chain by (1→6) glycosidic bond; →2)-ß-D-Galp-(1→ and α-L-Fucp-(1→ can also connect to →4)-ß-D-Xylp-(1→ forming di- or trisaccharide side chains.

Computed tomography-based COVID-19 triage through a deep neural network using mask-weighted global average pooling.

Background: There is an urgent need to find an effective and accurate method for triaging coronavirus disease 2019 (COVID-19) patients from millions or billions of people. Therefore, this study aimed to develop a novel deep-learning approach for COVID-19 triage based on chest computed tomography (CT) images, including normal, pneumonia, and COVID-19 cases. Methods: A total of 2,809 chest CT scans (1,105 COVID-19, 854 normal, and 850 non-3COVID-19 pneumonia cases) were acquired for this study and classified into the training set (n = 2,329) and test set (n = 480). A U-net-based convolutional neural network was used for lung segmentation, and a mask-weighted global average pooling (GAP) method was proposed for the deep neural network to improve the performance of COVID-19 classification between COVID-19 and normal or common pneumonia cases. Results: The results for lung segmentation reached a dice value of 96.5% on 30 independent CT scans. The performance of the mask-weighted GAP method achieved the COVID-19 triage with a sensitivity of 96.5% and specificity of 87.8% using the testing dataset. The mask-weighted GAP method demonstrated 0.9% and 2% improvements in sensitivity and specificity, respectively, compared with the normal GAP. In addition, fusion images between the CT images and the highlighted area from the deep learning model using the Grad-CAM method, indicating the lesion region detected using the deep learning method, were drawn and could also be confirmed by radiologists. Conclusions: This study proposed a mask-weighted GAP-based deep learning method and obtained promising results for COVID-19 triage based on chest CT images. Furthermore, it can be considered a convenient tool to assist doctors in diagnosing COVID-19.

Phytochemical constituents of Camellia osmantha fruit cores with antithrombotic activity.

Camellia osmantha is a new species of the genus Camellia and is an economically important ornamental plant. Its activity and ingredients are less studied than other Camellia plants. This study investigated the antithrombotic effect and chemical components of C. osmantha fruit cores using platelet aggregation assays and coagulation function tests. The cores of C. osmantha fruits were extracted with ethanol to obtain a crude extract. The extract was dissolved in water and further eluted with different concentrations of methanol on an MCI resin column to obtain three fractions. These samples were used for antithrombotic activity tests and phytochemical analysis. The results showed that the extract and its fractions of C. osmantha have strong antithrombotic activity, significantly reducing the platelet aggregation rate and prolonging the thrombin time (TT). The total saponins, flavonoids, and polyphenols in the active fractions may be responsible for the antithrombotic activity. The chemical constituents were analyzed by ultra-performance liquid chromatography-quadrupole time-of-flight mass spectrometry (UPLC-QTOF/MS). Twenty-three compounds were identified rapidly and accurately. Among them, ellagic acid, naringenin, and quercetin 3-O-glucuronide may be important antithrombotic constituents. Furthermore, interactions between these compounds and the P2Y1 receptor were investigated via molecular modeling, because the P2Y1 receptor is a key drug target of antiplatelet aggregative activity. The molecular docking results suggested that these compounds could combine tightly with the P2Y1R protein. Our results showed that C. osmantha fruit cores are rich in polyphenols, flavonoids, and saponins, which can be developed into a promising antithrombotic functional beverage for the prevention and treatment of cardiovascular and cerebrovascular diseases.

A Liver Segmentation Method Based on the Fusion of VNet and WGAN.

Accurate segmentation of liver images is an essential step in liver disease diagnosis, treatment planning, and prognosis. In recent years, although liver segmentation methods based on 2D convolutional neural networks have achieved good results, there is still a lack of interlayer information that causes severe loss of segmentation accuracy to a certain extent. Meanwhile, making the best of high-level and low-level features more effectively in a 2D segmentation network is a challenging problem. Therefore, we designed and implemented a 2.5-dimensional convolutional neural network, VNet_WGAN, to improve the accuracy of liver segmentation. First, we chose three adjacent layers of a liver model as the input of our network and adopted two convolution kernels in series connection, which can integrate cross-sectional spatial information and interlayer information of liver models. Second, a chain residual pooling module is added to fuse multilevel feature information to optimize the skip connection. Finally, the boundary loss function in the generator is employed to compensate for the lack of marginal pixel accuracy in the Dice loss function. The effectiveness of the proposed method is verified on two datasets, LiTS and CHAOS. The Dice coefficients are 92% and 90%, respectively, which are better than those of the compared segmentation networks. In addition, the experimental results also show that the proposed method can reduce computational consumption while retaining higher segmentation accuracy, which is significant for liver segmentation in practice and provides a favorable reference for clinicians in liver segmentation.

MTQA: Text-Based Multitype Question and Answer Reading Comprehension Model.

Text-based multitype question answering is one of the research hotspots in the field of reading comprehension models. Multitype reading comprehension models have the characteristics of shorter time to propose, complex components of relevant corpus, and greater difficulty in model construction. There are relatively few research works in this field. Therefore, it is urgent to improve the model performance. In this paper, a text-based multitype question and answer reading comprehension model (MTQA) is proposed. The model is based on a multilayer transformer encoding and decoding structure. In the decoding structure, the headers of the answer type prediction decoding, fragment decoding, arithmetic decoding, counting decoding, and negation are added for the characteristics of multiple types of corpora. Meanwhile, high-performance ELECTRA checkpoints are employed, and secondary pretraining based on these checkpoints and an absolute loss function are designed to improve the model performance. The experimental results show that the performance of the proposed model on the DROP and QUOREF corpora is better than the best results of the current existing models, which proves that the proposed MTQA model has high feature extraction and relatively strong generalization capabilities.

Identification and expression analysis of PUB genes in tea plant exposed to anthracnose pathogen and drought stresses.

The plant U-box (PUB) gene family, one of the major ubiquitin ligase families in plants, plays important roles in multiple cellular processes including environmental stress responses and resistance. The function of U-box genes has been well characterized in Arabidopsis and other plants. However, little is known about the tea plant (Camellia sinensis) PUB genes. Here, 89 U-box proteins were identified from the chromosome-scale referenced genome of tea plant. According to the domain organization and phylogenetic analysis, the tea plant PUB family were classified into ten classes, named Class I to X, respectively. Using previously released stress-related RNA-seq data in tea plant, we identified 34 stress-inducible CsPUB genes. Specifically, eight CsPUB genes were expressed differentially under both anthracnose pathogen and drought stresses. Moreover, six of the eight CsPUBs were upregulated in response to these two stresses. Expression profiling performed by qRT-PCR was consistent with the RNA-seq analysis, and stress-related cis-acting elements were identified in the promoter regions of the six upregulated CsPUB genes. These results strongly implied the putative functions of U-box ligase genes in response to biotic and abiotic stresses in tea plant.

The complete chloroplast genome sequence of Camellia rostrata S. X. Yang & S. F. Chai (Theaceae), a critically endangered yellow camellia from southwest China.

Camellia rostrata S. X. Yang & S. F. Chai is a recently described yellow camellia species from Guangxi, China. It is a critically endangered species according to the IUCN Red List Categories and Criteria. Here, we report the complete chloroplast (cp) genome based on next-generation sequencing technology. The complete cp genome of C. rostrata is 156,547 bp in length and consists of a large single-copy (LSC, 86,199 bp) region, a small single-copy (SSC, 18,204 bp) region, and a pair of inverted repeats (IRs, 26,072 bp). The genome contains 135 genes including 40 tRNA, eight rRNA, and 87 protein-coding genes. Phylogenetic analysis resolved C. rostrata in a clade containing C. huana and C. impressinervis, both of which are classified to Camellia sect. Archecamellia. Our findings support the placement of C. rostrata in C. sect. Archecamellia as proposed by a previous study. The cp genome of C. rostrata provides valuable bioinformatic resources for the protection and utilization of this yellow camellia species.

The complete chloroplast genome sequence of Camellia zhaiana (Theaceae), a critically endangered species from China.

Camellia zhaiana S.X. Yang (Theaceae) is a recently described species reported from Guangxi, China. It was proposed as a critically endangered species according to the IUCN Red List Categories and Criteria. In this study, we report and characterize the complete chloroplast (cp) genome of C. zhaiana using Illumina pair-end sequencing data. This is the first report of a cp genome of a species classified in Camellia section. Longipedicellata. The cp genome of C. zhaiana is 156,627 bp in length and includes a large single-copy region (LSC, 86,196 bp), a small single-copy region (SSC, 18,281 bp), and a pair of inverted repeat regions (IRs, 26,075 bp). The genome contains 135 genes, including 40 tRNA, eight rRNA, and 87 protein-coding genes. Phylogenetic analysis showed a strongly supported sister relationship between C. zhaiana and C. longipedicellata, which is a species classified in sect. Longipedicellata. These data support the previous systematic findings of C. zhaiana and advance the bioinformatics of the genus Camellia.

Insight into the roles of endogenous minerals in the activation of persulfate by graphitized biochar for tetracycline removal.

Owing to its environmental-friendliness, low-cost, and outstanding characteristics, biochar has been widely used for the catalytic degradation of various organic pollutants. In this study, a pre- and post-deashing graphitized biochar (DBC800 and PBC800-A) was prepared and compared with the pristine biochar (PBC800) to activate persulfate (PS) for tetracycline (TC) degradation. The influence of the natural endogenous mineral on the catalytic ability of biochar was investigated. Characterization results show that the inherent endogenous mineral in biochar not only acted as a natural pore-forming agent to promote the formation of the porous structure, but also facilitated the formation of edge defective structures, and altered the surface functional groups, as well as increased the carbonization and graphitization degree of biochar. The PBC800-A exhibited a much higher catalytic efficiency on PS activation and TC oxidative degradation with the reaction rate of 0.06055 min-1, 7.14 times as that of DBC800 (0.00861 min-1) and 4.63 times as that of PBC800 (0.00158 min-1). The endogenous minerals were conducive to the generation of free radicals and promoted the oxidative degradation of TC, which was mainly attributed to the improved carbon configuration. The post-deashing treatment was also found to significantly improve the electron transport efficiency of biochar by removing the residual ash, thereby promoting the generation of singlet oxygen. This study demonstrated that the natural minerals in biochar was beneficial for the degradation of TC, and more alternative natural minerals can be applied to co-pyrolysis with biochar for the remediation of refractory organic pollutants.

Feasibility of quantitative parameters of dynamically enhanced patterns of spiral computed tomography scanning integrated into tumour progression before targeted treatment of non-small cell lung cancer.

INTRODUCTION: The relationship between quantitative parameters of contrast-enhanced computed tomography (CT) and non-small cell lung cancer (NSCLC) progression remains controversial. We aimed to explore the usefulness of contrast-enhanced spiral CT scanning for confirming the time of tumour progression before targeted treatment of NSCLC. METHODS: Contrast-enhanced spiral CT scanning was performed on 33 NSCLC patients with a biopsy-proven diagnosis of NSCLC. All the patients were divided into three groups according to times of tumour progression (<6 weeks, 6-20 weeks, and >20 weeks). The perfusion CT data were used to calculate quantitative parameters, including enhanced peak values, peak time of tumour enhancement, ratio of tumour mass and enhanced aorta peak value and perfusion value of blood flow. Variance analysis was used for statistical analysis among the three groups using SAS 9.13 statistical software. RESULTS: Tumour perfusion values among the three group with different stage of TTP were significantly different from each other with P = 0.0129 (<6 weeks, perfusion value = 0.35 ± 0.15 mL/(min × mL); 6-20 weeks, perfusion value = 0.41 ± 0.086 mL/(min × mL); > 20 weeks, perfusion value = 0.47 ± 0.087 mL/(min × mL)). However, no significant differences were found in other parameters (enhanced peak values, peak time of tumour enhancement, ratios of tumour mass, and enhanced aorta peak value) among three groups (P > 0.05). CONCLUSION: The NSCLC patients with high perfusion value before targeted therapy are more sensitive to targeted therapy, and further experiments with larger sample size are needed.

A method of protein model classification and retrieval using bag-of-visual-features.

In this paper we propose a novel visual method for protein model classification and retrieval. Different from the conventional methods, the key idea of the proposed method is to extract image features of proteins and measure the visual similarity between proteins. Firstly, the multiview images are captured by vertices and planes of a given octahedron surrounding the protein. Secondly, the local features are extracted from each image of the different views by the SURF algorithm and are vector quantized into visual words using a visual codebook. Finally, KLD is employed to calculate the similarity distance between two feature vectors. Experimental results show that the proposed method has encouraging performances for protein retrieval and categorization as shown in the comparison with other methods.