In this paper, we designed an end-to-end spectral-spatial residual network (SSRN) that takes raw 3-D cubes as input data without feature engineering for hyperspectral image classification. In this ...network, the spectral and spatial residual blocks consecutively learn discriminative features from abundant spectral signatures and spatial contexts in hyperspectral imagery (HSI). The proposed SSRN is a supervised deep learning framework that alleviates the declining-accuracy phenomenon of other deep learning models. Specifically, the residual blocks connect every other 3-D convolutional layer through identity mapping, which facilitates the backpropagation of gradients. Furthermore, we impose batch normalization on every convolutional layer to regularize the learning process and improve the classification performance of trained models. Quantitative and qualitative results demonstrate that the SSRN achieved the state-of-the-art HSI classification accuracy in agricultural, rural-urban, and urban data sets: Indian Pines, Kennedy Space Center, and University of Pavia.
Versatile and precise genome modifications are needed to create a wider range of adoptive cellular therapies
. Here we report two improvements that increase the efficiency of CRISPR-Cas9-based genome ...editing in clinically relevant primary cell types. Truncated Cas9 target sequences (tCTSs) added at the ends of the homology-directed repair (HDR) template interact with Cas9 ribonucleoproteins (RNPs) to shuttle the template to the nucleus, enhancing HDR efficiency approximately two- to fourfold. Furthermore, stabilizing Cas9 RNPs into nanoparticles with polyglutamic acid further improves editing efficiency by approximately twofold, reduces toxicity, and enables lyophilized storage without loss of activity. Combining the two improvements increases gene targeting efficiency even at reduced HDR template doses, yielding approximately two to six times as many viable edited cells across multiple genomic loci in diverse cell types, such as bulk (CD3
) T cells, CD8
T cells, CD4
T cells, regulatory T cells (Tregs), γδ T cells, B cells, natural killer cells, and primary and induced pluripotent stem cell-derived
hematopoietic stem progenitor cells (HSPCs).
The diagnosis of postacute sequelae of coronavirus disease 2019 (PASC) poses an ongoing medical challenge. To identify biomarkers associated with PASC we analyzed plasma samples collected from PASC ...and coronavirus disease 2019 patients to quantify viral antigens and inflammatory markers. We detect severe acute respiratory syndrome coronavirus 2 spike predominantly in PASC patients up to 12 months after diagnosis.
Recently, the advancement of deep learning (DL) in discriminative feature learning from 3-D LiDAR data has led to rapid development in the field of autonomous driving. However, automated processing ...uneven, unstructured, noisy, and massive 3-D point clouds are a challenging and tedious task. In this article, we provide a systematic review of existing compelling DL architectures applied in LiDAR point clouds, detailing for specific tasks in autonomous driving, such as segmentation, detection, and classification. Although several published research articles focus on specific topics in computer vision for autonomous vehicles, to date, no general survey on DL applied in LiDAR point clouds for autonomous vehicles exists. Thus, the goal of this article is to narrow the gap in this topic. More than 140 key contributions in the recent five years are summarized in this survey, including the milestone 3-D deep architectures, the remarkable DL applications in 3-D semantic segmentation, object detection, and classification; specific data sets, evaluation metrics, and the state-of-the-art performance. Finally, we conclude the remaining challenges and future researches.
This study aims to provide a national benchmark rate of post-tonsillectomy haemorrhage (PTH) in Australia. Using data from Australia's National Hospital Morbidity Database (NHMD) from 1 July 2000 to ...30 June 2020, we have conducted a nation-wide population-based study to estimate a reference rate of PTH. Outcomes of interest included the overall rate and time-trend of PTH, the relationship between PTH rates with age and gender as well as the epidemiology of tonsillectomy procedures. A total of 941,557 tonsillectomy procedures and 15,391 PTH episodes were recorded for the study period. Whilst the incidence of tonsillectomy procedures and the number of day-stay tonsillectomy procedures have increased substantially over time, the overall rate of PTH for all ages has remained relatively constant (1.6% 95% CI: 1.61 to 1.66) with no significant association observed between the annual rates of PTH and time (year) (Spearman correlation coefficient, R.sub.s = 0.24 (95% CI: -0.22 to 0.61), P = 0.3). However, the rate of PTH in adults (aged 15 years and over) experienced a statistically significant mild to moderate upward association with time (year) R.sub.s = 0.64 (95% CI: 0.28 to 0.84), P = 0.003. Analysis of the odds of PTH using the risk factors of increasing age and male gender showed a unique age and gender risk pattern for PTH where males aged 20 to 24 years had the highest risk of PTH odds ratio 7.3 (95% CI: 6.7 to 7.8) compared to patients aged 1 to 4 years. Clinicians should be mindful of the greater risk of PTH in male adolescents and young adults. The NHMD datasets can be continually used to evaluate the benchmark PTH rate in Australia and to facilitate tonsillectomy surgical audit activities and quality improvement programs on a national basis.
Many individuals mount nearly identical antibody responses to SARS-CoV-2. To gain insight into how the viral spike (S) protein receptor-binding domain (RBD) might evolve in response to common ...antibody responses, we studied mutations occurring during virus evolution in a persistently infected immunocompromised individual. We use antibody Fab/RBD structures to predict, and pseudotypes to confirm, that mutations found in late-stage evolved S variants confer resistance to a common class of SARS-CoV-2 neutralizing antibodies we isolated from a healthy COVID-19 convalescent donor. Resistance extends to the polyclonal serum immunoglobulins of four out of four healthy convalescent donors we tested and to monoclonal antibodies in clinical use. We further show that affinity maturation is unimportant for wild-type virus neutralization but is critical to neutralization breadth. Because the mutations we studied foreshadowed emerging variants that are now circulating across the globe, our results have implications to the long-term efficacy of S-directed countermeasures.
Display omitted
•SARS-CoV-2 spike evolves during persistent infection to resist common antibodies•Antibody affinity maturation is critical to neutralization breadth•Intra-host evolution foreshadows mutations in circulating spike variants
Structural and functional analysis of the evolution of SARS-CoV-2 in a persistently infected immunocompromised individual yields insights into mutations that foreshadow emerging viral variants worldwide.
High throughput sequencing has previously identified differentially expressed genes (DEGs) and enriched signalling networks in human myometrium for term (≥37 weeks) gestation labour, when defined as ...a singular state of activity at comparison to the non-labouring state. However, transcriptome changes that occur during transition from early to established labour (defined as ≤3 and >3 cm cervical dilatation, respectively) and potentially altered by fetal membrane rupture (ROM), when adapting from onset to completion of childbirth, remained to be defined. In the present study, we assessed whether differences for these two clinically observable factors of labour are associated with different myometrial transcriptome profiles. Analysis of our tissue ('bulk') RNA-seq data (NCBI Gene Expression Omnibus: GSE80172) with classification of labour into four groups, each compared to the same non-labour group, identified more DEGs for early than established labour; ROM was the strongest up-regulator of DEGs. We propose that lower DEGs frequency for early labour and/or ROM negative myometrium was attributed to bulk RNA-seq limitations associated with tissue heterogeneity, as well as the possibility that processes other than gene transcription are of more importance at labour onset. Integrative analysis with future data from additional samples, which have at least equivalent refined clinical classification for labour status, and alternative omics approaches will help to explain what truly contributes to transcriptomic changes that are critical for labour onset. Lastly, we identified five DEGs common to all labour groupings; two of which (AREG and PER3) were validated by qPCR and not differentially expressed in placenta and choriodecidua.
Review: Deep Learning on 3D Point Clouds Bello, Saifullahi Aminu; Yu, Shangshu; Wang, Cheng ...
Remote sensing (Basel, Switzerland),
06/2020, Letnik:
12, Številka:
11
Journal Article
Recenzirano
Odprti dostop
A point cloud is a set of points defined in a 3D metric space. Point clouds have become one of the most significant data formats for 3D representation and are gaining increased popularity as a result ...of the increased availability of acquisition devices, as well as seeing increased application in areas such as robotics, autonomous driving, and augmented and virtual reality. Deep learning is now the most powerful tool for data processing in computer vision and is becoming the most preferred technique for tasks such as classification, segmentation, and detection. While deep learning techniques are mainly applied to data with a structured grid, the point cloud, on the other hand, is unstructured. The unstructuredness of point clouds makes the use of deep learning for its direct processing very challenging. This paper contains a review of the recent state-of-the-art deep learning techniques, mainly focusing on raw point cloud data. The initial work on deep learning directly with raw point cloud data did not model local regions; therefore, subsequent approaches model local regions through sampling and grouping. More recently, several approaches have been proposed that not only model the local regions but also explore the correlation between points in the local regions. From the survey, we conclude that approaches that model local regions and take into account the correlation between points in the local regions perform better. Contrary to existing reviews, this paper provides a general structure for learning with raw point clouds, and various methods were compared based on the general structure. This work also introduces the popular 3D point cloud benchmark datasets and discusses the application of deep learning in popular 3D vision tasks, including classification, segmentation, and detection.