Search

Banana phenotypic parameters are one of the important elements in the study of banana growth and development. Through the measurement of banana phenotypic parameters, we can obtain information about the growth status, nutritional status and quality indexes of banana plants, and pseudo-stem parameters are significant indicators in banana phenotypic parameters. This research proposes a two-stage approach combining morphological features and deep learning point cloud segmentation to extract banana pseudo-stem parameters. Specifically, in the first step, seed points are extracted using the DBSCAN clustering algorithm, and banana individual plant segmentation is accomplished using the region growing algorithm based on seed points. Its precision, recall and F1-score were 97.73%, 97.36% and 97.54%, respectively. This indicates that the DBSCAN clustering algorithm and the seed point based region growing algorithm can effectively realize the plant count of banana plants and initially realize the individual plant segmentation of banana. The second step is to use PointNet++, PointNet, and DGCNN for pseudo-stem and canopy segmentation of individual banana plants. All three models perform well in segmentation, with PointNet++ performing the best. Its precision, recall, F1-score, Matthews correlation coefficient and Dice coefficient reached 0.9956, 0.9709, 0.9831, 0.9670 and 0.9831. This shows that deep learning has a better applicability in segmenting banana plants. In the results of segmentation, we measure the banana pseudo-stem circumference and pseudo-stem height. The correlation between the extracted pseudo-stem height and pseudo-stem circumference compared to the measured values was 96.70% and 82.32%, respectively. The above two-stage method of extracting banana pseudo-stem parameters overcomes the difficulties of point cloud individual plant segmentation associated with intensive banana cultivation. It makes the management of individual banana plants possible and provides accurate phenotypic parameter information for banana plantation management. It lays the foundation for further assessment of banana growth and nutritional status.

Non-small cell lung cancer (NSCLC) has a high morbidity and mortality, and it is imperative to explore the latent pathogenesis mechanism of NSCLC progression to find potential prognostic biomarkers and therapeutic targets. The present study aimed to explore the biological function of circSHKBP1 in NSCLC. circSHKBP1 was found to be upregulated in NSCLC tissues and cell lines and was enriched in exosomes derived from NSCLC cells. Exosomal circSHKBP1 enhanced the proliferation, migration, invasion, and stemness of NSCLC cells. miRNA-1294 was identified as a target for circSHKBP1, and circSHKBP1 upregulated PKM2 expression by sponging miR-1294. Exosomal circSHKBP1 regulated glycolysis through PKM2 in a HIF-1α-dependent manner in NSCLC cells and promoted M2 polarization and macrophage recruitment. Moreover, exosomal circSHKBP1 promoted NSCLC cell growth, metastasis, and M2 infiltration in vivo. Thus, exosomal circSHKBP1 participated in the progression of NSCLC via the miR-1294/PKM2 axis. circSHKBP1 may be potential biomarker for the diagnosis and treatment of NSCLC.

Abstract High heterogeneity of lung adenocarcinoma (LUAD) is a major clinical challenge. This study aims to characterize the molecular features of LUAD through classification based on metabolism-related genes. A total of 500 LUAD samples from The Cancer Genome Atlas (TCGA) and 612 from Gene Expression Omnibus (GEO) were integrated with 2,753 metabolism-related genes to determine the molecular classification. Systematic bioinformatics analysis was used to conduct correlation analysis between metabolism-related classification and molecular characteristics of LUAD. LUAD patients were divided into three molecular clusters (C1-C3). Survival analysis revealed that C1 and C2 showed good and poor prognoses, respectively. Associational analysis of classification and molecular characteristics revealed that C1 was associated with low pathological stage, metabolic pathways, high metabolic process, active immune process and checkpoint, sensitive drug response, as well as a low genetic mutation. Nevertheless, C2 was associated with high pathological stage, carcinogenic pathways, low metabolic process, inactive immune signatures, resistant drug response, and frequent genetic mutation. Eventually, a classifier with 60 metabolic genes was constructed, confirming the robustness of molecular classification on LUAD. Our findings promote the understanding of LUAD molecular characteristics, and the research data may be used for providing information be helpful for clinical diagnosis and treatment.

Visual Question Answering (VQA) about diseases is an essential feature of intelligent management in smart agriculture. Currently, research on fruit tree diseases using deep learning mainly uses single-source data information, such as visible images or spectral data, yielding classification and identification results that cannot be directly used in practical agricultural decision-making. In this study, a VQA model for fruit tree diseases based on multimodal feature fusion was designed. Fusing images and Q&A knowledge of disease management, the model obtains the decision-making answer by querying questions about fruit tree disease images to find relevant disease image regions. The main contributions of this study were as follows: (1) a multimodal bilinear factorized pooling model using Tucker decomposition was proposed to fuse the image features with question features: (2) a deep modular co-attention architecture was explored to simultaneously learn the image and question attention to obtain richer graphical features and interactivity. The experiments showed that the proposed unified model combining the bilinear model and co-attentive learning in a new network architecture obtained 86.36% accuracy in decision-making under the condition of limited data (8,450 images and 4,560k Q&A pairs of data), outperforming existing multimodal methods. The data augmentation is adopted on the training set to avoid overfitting. Ten runs of 10-fold cross-validation are used to report the unbiased performance. The proposed multimodal fusion model achieved friendly interaction and fine-grained identification and decision-making performance. Thus, the model can be widely deployed in intelligent agriculture.

Citrus psyllid is the only insect vector of citrus Huanglongbing (HLB), which is the most destructive disease in the citrus industry. There is no effective treatment for HLB, so detecting citrus psyllids as soon as possible is the key prevention measure for citrus HLB. It is time-consuming and laborious to search for citrus psyllids through artificial patrol, which is inconvenient for the management of citrus orchards. With the development of artificial intelligence technology, a computer vision method instead of the artificial patrol can be adopted for orchard management to reduce the cost and time. The citrus psyllid is small in shape and gray in color, similar to the stem, stump, and withered part of the leaves, leading to difficulty for the traditional target detection algorithm to achieve a good recognition effect. In this work, in order to make the model have good generalization ability under outdoor light condition, a high-definition camera to collect data set of citrus psyllids and citrus fruit flies under natural light condition was used, a method to increase the number of small target pests in citrus based on semantic segmentation algorithm was proposed, and the cascade region-based convolution neural networks (R-CNN) (convolutional neural network) algorithm was improved to enhance the recognition effect of small target pests using multiscale training, combining CBAM attention mechanism with high-resolution feature retention network high-resoultion network (HRNet) as feature extraction network, adding sawtooth atrous spatial pyramid pooling (ASPP) structure to fully extract high-resolution features from different scales, and adding feature pyramid networks (FPN) structure for feature fusion at different scales. To mine difficult samples more deeply, an online hard sample mining strategy was adopted in the process of model sampling. The results show that the improved cascade R-CNN algorithm after training has an average recognition accuracy of 88.78% for citrus psyllids. Compared with VGG16, ResNet50, and other common networks, the improved small target recognition algorithm obtains the highest recognition performance. Experimental results also show that the improved cascade R-CNN algorithm not only performs well in citrus psylla identification but also in other small targets such as citrus fruit flies, which makes it possible and feasible to detect small target pests with a field high-definition camera.

Using unmanned aerial vehicle (UAV) real-time remote sensing to monitor diseased plants or abnormal areas of orchards from a low altitude perspective can greatly improve the efficiency and response speed of the patrol in smart orchards. The purpose of this paper is to realize the intelligence of the UAV terminal and make the UAV patrol orchard in real-time. The existing lightweight object detection algorithms are usually difficult to consider both detection accuracy and processing speed. In this study, a new lightweight model named Swin-T YOLOX, which consists of the advanced detection network YOLOX and the strong backbone Swin Transformer, was proposed. Model layer pruning technology was adopted to prune the multi-layer stacked structure of the Swin Transformer. A variety of data enhancement strategies were conducted to expand the dataset in the model training stage. The lightweight Swin-T YOLOX model was deployed to the embedded platform Jetson Xavier NX to evaluate its detection capability and real-time performance of the UAV patrol mission in the orchard. The research results show that, with the help of TensorRT optimization, the proposed lightweight Swin-T YOLOX network achieved 94.0% accuracy and achieved a detection speed of 40 fps on the embedded platform (Jetson Xavier NX) for patrol orchard missions. Compared to the original YOLOX network, the model accuracy has increased by 1.9%. Compared to the original Swin-T YOLOX, the size of the proposed lightweight Swin-T YOLOX has been reduced to two-thirds, while the model accuracy has slightly increased by 0.7%. At the same time, the detection speed of the model has reached 40 fps, which can be applied to the real-time UAV patrol in the orchard.

Candida albicans is the most common fungal pathogen and relies on the Hog1-MAPK pathway to resist osmotic stress posed by the environment or during host invasions. Here, we investigated the role of SPT20 in response to osmotic stress. Testing a C. albicans spt20Δ/Δ mutant, we found it was sensitive to osmotic stress. Using sequence alignment, we identified the conserved functional domains between CaSpt20 and ScSpt20. Reconstitution of the Spt20 function in a spt20Δ/CaSPT20 complemented strain found CaSPT20 can suppress the high sensitivity to hyperosmotic stressors, a cell wall stress agent, and antifungal drugs in the Saccharomyces cerevisiae spt20Δ/Δ mutant background. We measured the cellular glycerol accumulation and found it was significantly lower in the C. albicans spt20Δ/Δ mutant strain, compared to the wild type strain SC5314 (P < 0.001). This result was also supported by quantitative reverse transcription-PCR, which showed the expression levels of gene contributing to glycerol accumulation were reduced in Caspt20Δ/Δ compared to wild type (GPD2 and TGL1, P < 0.001), while ADH7 and AGP2, whose expression can lead to glycerol decrease, were induced when cells were exposed to high osmolarity (ADH7, P < 0.001; AGP2, P = 0.002). In addition, we tested the transcription levels of Hog1-dependent osmotic stress response genes, and found that they were significantly upregulated in wild type cells encountering hyperosmolarity, while the expression of HGT10, SKO1, CAT1, and SLP3 were not induced when SPT20 was deleted. Although the transcript of ORF19.3661 and ORF19.4370 in Caspt20Δ/Δ was induced in the presence of 1 M NaCl, the levels were less than what was observed in the wild type (ORF19.3661, P = 0.007; ORF19.4370, P = 0.011). Moreover, the deletion of CaSPT20 in C. albicans reduced phosphorylation levels of Hog1. These findings suggested that SPT20 is conserved between yeast and C. albicans and plays an important role in adapting to osmotic stress through regulating Hog1-MAPK pathway.

Carbon dots (CDs) can be prepared from various organic (abundant) compounds that are rich in surfaces with –OH, –COOH, and –NH2 groups. Therefore, CDs exhibit good biocompatibility and electron transfer ability, allowing flexible surface modification and accelerated electron transfer during catalysis. Herein, CDs were prepared using a hydrothermal method with fructose, saccharose, and citric acid as C sources and urea as an N dopant. The as-prepared CDs were used to catalyze AgNO3–trisodium citrate (TSC) to produce Ag nanoparticles (AgNPs). The surface-enhanced Raman scattering (SERS) intensity increased with the increasing CDs concentration with Victoria blue B (VBB) as a signal molecule. The CDs exhibited a strong catalytic activity, with the highest activity shown by fructose-based CDs. After N doping, catalytic performance improved; with the passivation of a wrapped aptamer, the electron transfer was effectively disrupted (retarded). This resulted in the inhibition of the reaction and a decrease in the SERS intensity. When bisphenol A (BPA) was added, it specifically bound to the aptamer and CDs were released, recovering catalytical activity. The SERS intensity increased with BPA over the concentration range of 0.33–66.67 nmol/L. Thus, the aptamer-adjusted nanocatalytic SERS method can be applied for BPA detection.

We report a promising method for oxide-layer-free germanium (Ge)/silicon (Si) wafer bonding based on an amorphous Ge (a-Ge) intermediate layer between Si and Ge wafers. The effect of the exposure time (te), during which the a-Ge is exposed to the air after sputtering and being taken out of the chamber on the bubble density at the bonded interface, is identified and a near-bubble-free Ge/Si bonded interface is achieved for the te of 3 s. The crystallization of a-Ge at Ge/Si bonded interface starts from a-Ge/Ge interface and it fully turns to be single-crystal Ge after post-annealing. The oxide layer at a-Ge/a-Ge bonded interface formed by the interface hydrophilic reaction disappears due to the atom redistribution triggered by the crystallization of a-Ge. As expected, the performance of the Ge/Si heterojunction diode is significantly improved by this oxide-layer-free Ge/Si bonded interface. A low dark current of 1.6 µA, high on/off current ratio of 3.4  ×  105, and low ideality factor of 1.02 (150 K) is achieved at  −0.5 V for the bonded Ge/Si diode. Finally, the carrier transport mechanisms at Ge/Si bonded interface annealed at different temperatures are also clearly clarified. [ABSTRACT FROM AUTHOR]

The dependence of the morphology and crystallinity of an amorphous Ge (a-Ge) interlayer between two Si wafers on the annealing temperature is identified to understand the bubble evolution mechanism. The effect of a-Ge layer thickness on the bubble density and size at different annealing temperatures is also clearly clarified. It suggests that the bubble density is significantly affected by the crystallinity and thickness of the a-Ge layer. With the increase of the crystallinity and thickness of the a-Ge layer, the bubble density decreases. It is important that a near-bubble-free Ge interface, which is also an oxide-free interface, is achieved when the bonded Si wafers (a-Ge layer thickness  ⩾  20 nm) are annealed at 400 °C. Furthermore, the crystallization temperature of the a-Ge between the bonded Si wafers is lower than that on a Si substrate alone and the Ge grains firstly form at the Ge/Ge bonded interface, rather than the Ge/Si interface. We believe that the stress-induced crystallization of a-Ge film and the intermixing of Ge atoms at the Ge/Ge interface can be responsible for this feature. [ABSTRACT FROM AUTHOR]