RESUMEN
To accurately segment various clinical lesions from computed tomography(CT) images is a critical task for the diagnosis and treatment of many diseases. However, current segmentation frameworks are tailored to specific diseases, and limited frameworks can detect and segment different types of lesions. Besides, it is another challenging problem for current segmentation frameworks to segment visually inconspicuous and small-scale tumors (such as small intestinal stromal tumors and pancreatic tumors). Our proposed framework, CDI-NSTSEG, efficiently segments small non-salient tumors using multi-scale visual information and non-local target mining. CDI-NSTSEG follows the diagnostic process of clinicians, including preliminary screening, localization, refinement, and segmentation. Specifically, we first explore to extract the unique features at three different scales (1×, 0.5×, and 1.5×) based on the scale space theory. Our proposed scale fusion module (SFM) hierarchically fuses features to obtain a comprehensive representation, similar to preliminary screening in clinical diagnosis. The global localization module (GLM) is designed with a non-local attention mechanism. It captures the long-range semantic dependencies of channels and spatial locations from the fused features. GLM enables us to locate the tumor from a global perspective and output the initial prediction results. Finally, we design the layer focusing module (LFM) to gradually refine the initial results. LFM mainly conducts context exploration based on foreground and background features, focuses on suspicious areas layer-by-layer, and performs element-by-element addition and subtraction to eliminate errors. Our framework achieves state-of-the-art segmentation performance on small intestinal stromal tumor and pancreatic tumor datasets. CDI-NSTSEG outperforms the best comparison segmentation method by 7.38% Dice on small intestinal stromal tumors.
RESUMEN
BACKGROUND: Oral mucosal diseases are similar to the surrounding normal tissues, i.e., their many non-salient features, which poses a challenge for accurate segmentation lesions. Additionally, high-precision large models generate too many parameters, which puts pressure on storage and makes it difficult to deploy on portable devices. METHODS: To address these issues, we design a non-salient target segmentation model (NTSM) to improve segmentation performance while reducing the number of parameters. The NTSM includes a difference association (DA) module and multiple feature hierarchy pyramid attention (FHPA) modules. The DA module enhances feature differences at different levels to learn local context information and extend the segmentation mask to potentially similar areas. It also learns logical semantic relationship information through different receptive fields to determine the actual lesions and further elevates the segmentation performance of non-salient lesions. The FHPA module extracts pathological information from different views by performing the hadamard product attention (HPA) operation on input features, which reduces the number of parameters. RESULTS: The experimental results on the oral mucosal diseases (OMD) dataset and international skin imaging collaboration (ISIC) dataset demonstrate that our model outperforms existing state-of-the-art methods. Compared with the nnU-Net backbone, our model has 43.20% fewer parameters while still achieving a 3.14% increase in the Dice score. CONCLUSIONS: Our model has high segmentation accuracy on non-salient areas of oral mucosal diseases and can effectively reduce resource consumption.
Asunto(s)
Enfermedades de la Boca , Mucosa Bucal , Humanos , Enfermedades de la Boca/diagnóstico por imagen , Mucosa Bucal/patología , Mucosa Bucal/diagnóstico por imagen , Procesamiento de Imagen Asistido por Computador/métodosRESUMEN
WG-5 is a lightweight stream cipher proposed for usage in the resource-constrained devices, e.g., passive RFID tags, industrial controllers, contactless smart cards and sensors. In this paper, a weakness called slide property of WG-5 which has not been discovered in previous works is for the first time explored and analyzed. The result shows that the probability that two related key-IV pairs of WG-5 generate the shifted keystreams can be up to 2-20, which is significantly high compared with an ideal stream cipher that generates the random keystreams. The correctness and accuracy of this theoretical probability is confirmed experimentally. Based on the slide property of WG-5, some key recovery attacks on WG-5 in the related key setting are proposed. The cryptanalytic result shows that the 80-bit secret key of WG-5 can be recovered with a time complexity of 225.615, requiring 6 related keys and 80 keystream bits for each of 224.585 chosen IVs. The experimental result validates our attack and shows that WG-5 can be broken within about 92.054 seconds on a common PC in the related key setting. These results imply that the design of WG-5 is far from optimal and needs to be strengthened to provide enough security for the lightweight constrained applications.
RESUMEN
Accurate target segmentation from computed tomography (CT) scans is crucial for surgical robots to perform clinical surgeries successfully. However, the lack of medical image data and annotations has been the biggest obstacle to learning robust medical image segmentation models. Self-supervised learning can effectively address this problem by providing a strategy to pre-train a model with unlabeled data, and then fine-tune downstream tasks with limited labeled data. Existing self-supervised methods fail to simultaneously utilize the abundant global anatomical structure information and local feature differences in medical imaging. In this work, we propose a new strategy for the pre-training framework, which uses the three-dimensional anatomical structure of medical images and specific task and background cues to segment volumetric medical images with limited annotations. Specifically, we propose (1) learning intrinsic patterns of volumetric medical image structures through multiple sub-tasks, and (2) designing a multi-level background cube contrastive learning strategy to enhance the target feature representation by exploiting the differences between the specific target and background. We conduct extensive evaluations on two publicly available datasets. Under limited annotation settings, the proposed method yields significant improvements compared to other self-supervised learning techniques. The proposed method achieves within 6% of the baseline performance using only five labeled CT volumes for training. Once the paper is online, the code and dataset will be available at https://github.com/PinkGhost0812/SGL.
RESUMEN
The deposition of amyloid ß peptide (Aß) is one of the main pathological features of AD. The much-talked sensory gamma entrainment may be a new treatment for Aß load. Here we reviewed the generation and clearance pathways of Aß, aberrant gamma oscillation in AD, and the therapeutic effect of sensory gamma entrainment on AD. In addition, we discuss these results based on stimulus parameters and possible potential mechanisms. This provides the support for sensory gamma entrainment targeting Aß to improve AD.
Asunto(s)
Enfermedad de Alzheimer , Péptidos beta-Amiloides , Humanos , Péptidos beta-Amiloides/metabolismo , Enfermedad de Alzheimer/metabolismo , Proteínas Amiloidogénicas/uso terapéutico , Precursor de Proteína beta-Amiloide/metabolismo , Secretasas de la Proteína Precursora del Amiloide/metabolismoRESUMEN
The quality and safety of sufu fermented using Mucor racemosa M2 was studied and compared with naturally fermented sufu. After 90 days post-fermentation, both naturally fermented and inoculated fermented sufu reached the maturity standard of sufu, and the degree of protein hydrolysis of natural sufu (WP/TP: 34% ± 1%; AAN/TN: 33% ± 1%) was slightly higher than that of the inoculated sufu (WP/TP: 28.2% ± 0.4%; AAN/TN: 27% ± 1%). The hardness and adhesiveness of inoculated sufu (Hadness: 1063 g ± 211 g; Adhesiveness: -80 g ± 47 g) were significantly greater than those of natural sufu (Hadness: 790 g ± 57 g; Adhesiveness: -23 g ± 28 g), and the internal structure of natural sufu was denser and more uniform than that of inoculated sufu. A total of 50 aroma compounds were detected in natural and inoculated sufu. The total number of bacterial colonies in naturally fermented sufu was significantly higher than that in inoculated sufu, and the pathogenic bacteria in both types of fermented sufu were lower than the limit of pathogenic bacteria required in fermented soybean products. The content of biogenic amines in sufu was determined by high performance liquid chromatography (HPLC), and the results showed that the content of biogenic amines (Putrescine, Cadaverine, Histamine, Tyramine, etc.) in naturally fermented sufu was significantly higher than that in inoculated fermented sufu. Especially the histamine content, after 90 days of fermentation, was found to be 64.95 ± 4.55 for inoculated fertilization and 44.24 ± 0.71 for natural fertilization. Overall, the quality of inoculated sufu was somewhat better than that of natural sufu, and the M2 strain can be used to ferment sufu.
RESUMEN
This study aimed to investigate the properties of acidic whey tofu gelatin generated from two acidic whey coagulants by pure fermentation of Lactiplantibacillus paracasei and L. plantarum, as well as the characteristics of acidic whey tofu. The optimal holding temperature and the amount of coagulants added were determined based on the pH, water-holding capacity, texture, microstructure, and rheological properties of tofu gelation. Then, the differences in quality between tofu produced by pure bacterial fermentation and by natural fermentation were investigated under optimal tofu gelatin preparation conditions. The tofu gelatin presented the best texture at 37 °C with a 10% addition of coagulants fermented by both L. paracasei and L. plantarum. Under these conditions, the coagulant produced by the fermentation of L. plantarum resulted in a shorter formation time and stronger tofu gelatin compared with that produced from L. paracasei. Tofu produced by the fermentation of L. paracasei had higher pH, less hardness, and a rougher network structure, whereas tofu produced by the fermentation of L. plantarum was closer to tofu produced by natural fermentation in terms of pH, texture, rheology, and microstructure.
RESUMEN
We are concerned with using user-tagged images to learn proper hashing functions for image retrieval. The benefits are two-fold: (1) we could obtain abundant training data for deep hashing models; (2) tagging data possesses richer semantic information which could help better characterize similarity relationships between images. However, tagging data suffers from noises, vagueness and incompleteness. Different from previous unsupervised or supervised hashing learning, we propose a novel weakly-supervised deep hashing framework which consists of two stages: weakly-supervised pre-training and supervised fine-tuning. The second stage is as usual. In the first stage, we propose two formulations Tag-basEd weakLy-supErvised Modally COoperative hashing Network (TelecomNet) and Generalized TelecomNet (GTelecomNet). Rather than performing supervision on tags, TelecomNet first learns an observed semantic embedding vector for each image from attached tags and then uses it to guide hashing learning. GTelecomNet introduces a novel semantic network to exploit more precise semantic information. By carefully designing the optimization problem, they can well leverage tagging information and image content for hashing learning. The framework is general and does not depend on specific deep hashing methods. Empirical results on real world datasets show that they significantly increase the performance of state-of-the-art deep hashing methods.
RESUMEN
Multiview clustering aims to leverage information from multiple views to improve the clustering performance. Most previous works assumed that each view has complete data. However, in real-world datasets, it is often the case that a view may contain some missing data, resulting in the problem of incomplete multiview clustering (IMC). Previous approaches to this problem have at least one of the following drawbacks: 1) employing shallow models, which cannot well handle the dependence and discrepancy among different views; 2) ignoring the hidden information of the missing data; and 3) being dedicated to the two-view case. To eliminate all these drawbacks, in this work, we present the adversarial IMC (AIMC) framework. In particular, AIMC seeks the common latent representation of multiview data for reconstructing raw data and inferring missing data. The elementwise reconstruction and the generative adversarial network are integrated to evaluate the reconstruction. They aim to capture the overall structure and get a deeper semantic understanding, respectively. Moreover, the clustering loss is designed to obtain a better clustering structure. We explore two variants of AIMC, namely: 1) autoencoder-based AIMC (AAIMC) and 2) generalized AIMC (GAIMC), with different strategies to obtain the multiview common representation. Experiments conducted on six real-world datasets show that AAIMC and GAIMC perform well and outperform the baseline methods.
Asunto(s)
Algoritmos , Semántica , Análisis por ConglomeradosRESUMEN
Multi-keyword query is widely supported in text search engines. However, an analogue in image retrieval systems, multi-object query, is rarely studied. Meanwhile, traditional object-based image retrieval methods often involve multiple steps separately. In this work, we propose a weakly-supervised Deep Multiple Instance Hashing (DMIH) approach for multi-object image retrieval. Our DMIH approach, which leverages a popular CNN model to build the end-to-end relation between a raw image and the binary hash codes of its multiple objects, can support multi-object queries effectively and integrate object detection with hashing learning seamlessly. We treat object detection as a binary multiple instance learning (MIL) problem and such instances are automatically extracted from multi-scale convolutional feature maps. We also design a conditional random field (CRF) module to capture both the semantic and spatial relations among different class labels. For hashing training, we sample image pairs to learn their semantic relationships in terms of hash codes of the most probable proposals for owned labels as guided by object predictors. The two objectives benefit each other in a multi-task learning scheme. Finally, a two-level inverted index method is proposed to further speed up the retrieval of multi-object queries. Our DMIH approach outperforms state-of-the-arts on public benchmarks for object-based image retrieval and achieves promising results for multi-object queries.
RESUMEN
Although the strategies used by bacteria to adapt to specific environmental conditions are widely reported, fewer studies have addressed how microbes with a cosmopolitan distribution can survive in diverse ecosystems. Exiguobacterium is a versatile genus whose members are commonly found in various habitats. To better understand the mechanisms underlying the universality of Exiguobacterium, we collected 105 strains from diverse environments and performed large-scale metabolic and adaptive ability tests. We found that most Exiguobacterium members have the capacity to survive under wide ranges of temperature, salinity, and pH. According to phylogenetic and average nucleotide identity analyses, we identified 27 putative species and classified two genetic groups: groups I and II. Comparative genomic analysis revealed that the Exiguobacterium members utilize a variety of complex polysaccharides and proteins to support survival in diverse environments and also employ a number of chaperonins and transporters for this purpose. We observed that the group I species can be found in more diverse terrestrial environments and have a larger genome size than the group II species. Our analyses revealed that the expansion of transporter families drove genomic expansion in group I strains, and we identified 25 transporter families, many of which are involved in the transport of important substrates and resistance to environmental stresses and are enriched in group I strains. This study provides important insights into both the overall general genetic basis for the cosmopolitan distribution of a bacterial genus and the evolutionary and adaptive strategies of Exiguobacterium. IMPORTANCE The wide distribution characteristics of Exiguobacterium make it a valuable model for studying the adaptive strategies of bacteria that can survive in multiple habitats. In this study, we reveal that members of the Exiguobacterium genus have a cosmopolitan distribution and share an extensive adaptability that enables them to survive in various environments. The capacities shared by Exiguobacterium members, such as their diverse means of polysaccharide utilization and environmental-stress resistance, provide an important basis for their cosmopolitan distribution. Furthermore, the selective expansion of transporter families has been a main driving force for genomic evolution in Exiguobacterium. Our findings improve our understanding of the adaptive and evolutionary mechanisms of cosmopolitan bacteria and the vital genomic traits that can facilitate niche adaptation.
RESUMEN
Multiview representation learning (MVRL) leverages information from multiple views to obtain a common representation summarizing the consistency and complementarity in multiview data. Most previous matrix factorization-based MVRL methods are shallow models that neglect the complex hierarchical information. The recently proposed deep multiview factorization models cannot explicitly capture consistency and complementarity in multiview data. We present the deep multiview concept learning (DMCL) method, which hierarchically factorizes the multiview data, and tries to explicitly model consistent and complementary information and capture semantic structures at the highest abstraction level. We explore two variants of the DMCL framework, DMCL-L and DMCL-N, with respectively linear/nonlinear transformations between adjacent layers. We propose two block coordinate descent-based optimization methods for DMCL-L and DMCL-N. We verify the effectiveness of DMCL on three real-world data sets for both clustering and classification tasks.
RESUMEN
This paper addresses the task of query-focused video summarization, which takes user queries and long videos as inputs and generates query-focused video summaries. Compared to video summarization, which mainly concentrates on finding the most diverse and representative visual contents as a summary, the task of query-focused video summarization considers the user's intent and the semantic meaning of generated summary. In this paper, we propose a method, named query-biased self-attentive network (QSAN) to tackle this challenge. Our key idea is to utilize the semantic information from video descriptions to generate a generic summary and then to combine the information from the query to generate a query-focused summary. Specifically, we first propose a hierarchical self-attentive network to model the relative relationship at three levels, which are different frames from a segment, different segments of the same video, textual information of video description and its related visual contents. We train the model on video caption dataset and employ a reinforced caption generator to generate a video description, which can help us locate important frames or shots. Then we build a query-aware scoring module to compute the query-relevant score for each shot and generate the query-focused summary. Extensive experiments on the benchmark dataset demonstrate the competitive performance of our approach compared to some methods.
RESUMEN
Nowadays, a lot of people possess accounts on multiple online social networks, e.g., Facebook and Twitter. These networks are overlapped, but the correspondences between their users are not explicitly given. Mapping common users across these social networks will be beneficial for applications such as cross-network recommendation. In recent years, a lot of mapping algorithms have been proposed which exploited social and/or profile relations between users from different networks. However, there is still a lack of unified mapping framework which can well exploit high-order relational information in both social structures and profiles. In this paper, we propose a unified hypergraph learning framework named unified manifold alignment on hypergraph (UMAH) for this task. UMAH models social structures and user profile relations in a unified hypergraph where the relative weights of profile hyperedges are determined automatically. Given a set of training user correspondences, a common subspace is learned by preserving the hypergraph structure as well as the correspondence relations of labeled users. UMAH intrinsically performs semisupervised manifold alignment with profile information for calibration. For a target user in one network, UMAH ranks all the users in the other network by their probabilities of being the corresponding user (measured by similarity in the subspace). In experiments, we evaluate UMAH on three real world data sets and compare it to state-of-art baseline methods. Experimental results have demonstrated the effectiveness of UMAH in mapping users across networks.
RESUMEN
Root-knot nematodes (RKNs) can infect almost all crops, and result in huge economic losses in agriculture. There is no effective and environmentally safe means available to control RKNs. Alcaligenes faecalis ZD02 isolated from free living nematode Caenorhabditis elegans cadavers shows toxicity against RKN Meloidogyne incognita, that makes this strain to be a good bionematicide candidate for controlling of RKNs. Here, we firstly report the complete genome of A. faecalis ZD02 and describe its features. Additionally, we found two potential virulence factors in this genome, which play important roles for the nematocidal activity of A. faecalis ZD02.
Asunto(s)
Alcaligenes faecalis/genética , Genoma Bacteriano , Alcaligenes faecalis/química , Alcaligenes faecalis/aislamiento & purificación , Animales , Antinematodos/aislamiento & purificación , Agentes de Control Biológico/aislamiento & purificación , Caenorhabditis elegans , Tylenchoidea , Factores de Virulencia/aislamiento & purificaciónRESUMEN
BACKGROUND: Prokaryotic plasmids have played significant roles in the evolution of bacterial genomes and have a great impact on the metabolic functions of the host cell. Many bacterial strains contain multiple plasmids, but the relationships between bacterial plasmids and chromosomes are unclear. We focused on plasmids from the Bacillus cereus group because most strains contain several plasmids. RESULTS: We collected the genome sequences of 104 plasmids and 20 chromosomes from B. cereus group strains, and we studied the relationships between plasmids and chromosomes by focusing on the pan-genomes of these plasmids and chromosomes. In terms of basic features (base composition and codon usage), the genes on plasmids were more similar to the chromosomal variable genes (distributed genes and unique genes) than to the chromosomal core genes. Although all the functional categories of the chromosomal genes were exhibited by the plasmid genes, the proportions of each category differed between these two gene sets. The 598 gene families shared between chromosomes and plasmids displayed a uniform distribution between the two groups. A phylogenetic analysis of the shared genes, including the chromosomal core gene set, indicated that gene exchange events between plasmids and chromosomes occurred frequently during the evolutionary histories of the strains and species in this group. Moreover, the shared genes between plasmids and chromosomes usually had different promoter and terminator sequences, suggesting that they are regulated by different elements at the transcriptional level. CONCLUSIONS: We speculate that for the entire B. cereus group, adaptive genes are preserved on both plasmids and chromosomes; however, in a single cell, homologous genes on plasmids and the chromosome are controlled by different regulators to reduce the burden of maintaining redundant genes.