scholarly journals ImPartial: Partial Annotations for Cell Instance Segmentation

2021 ◽  
Author(s):  
Natalia Martinez ◽  
Guillermo Sapiro ◽  
Allen Tannenbaum ◽  
Travis J. Hollmann ◽  
Saad Nadeem
Keyword(s):  
Supervised Classification ◽  
Superior Performance ◽  
Imaging Modalities ◽  
Clinical Settings ◽  
Mixture Of Distributions ◽  
Parallel Task ◽  
Time Costs ◽  
Link Type ◽  
Multiplexed Imaging ◽  
Instance Segmentation

Segmenting noisy multiplex spatial tissue images constitutes a challenging task, since the characteristics of both the noise and the biology being imaged differs significantly across tissues and modalities; this is compounded by the high monetary and time costs associated with manual annotations. It is therefore imperative to build algorithms that can accurately segment the noisy images based on a small number of annotations. Recently techniques to derive such an algorithm from a few scribbled annotations have been proposed, mostly relying on the refinement and estimation of pseudo-labels. Other techniques leverage the success of self-supervised denoising as a parallel task to potentially improve the segmentation objective when few annotations are available. In this paper, we propose a method that augments the segmentation objective via self-supervised multi-channel quantized imputation, meaning that each class of the segmentation objective can be characterized by a mixture of distributions. This approach leverages the observation that perfect pixel-wise reconstruction or denoising of the image is not needed for accurate segmentation, and introduces a self-supervised classification objective that better aligns with the overall segmentation goal. We demonstrate the superior performance of our approach for a variety of cancer datasets acquired with different highly-multiplexed imaging modalities in real clinical settings. Code for our method along with a benchmarking dataset is available at https://github.com/natalialmg/ImPartial.

Assessing the Unacquainted: Inferred Reviewer Personality and Review Helpfulness

MIS Quarterly ◽  
2021 ◽  
Vol 45 (3) ◽  
pp. 1113-1148
Author(s):  
Angela Xia Liu ◽  
◽  
Yilin Li ◽  
Sean Xu ◽  
◽  
...  
Keyword(s):  
Personality Traits ◽  
Supervised Classification ◽  
Large Population ◽  
Superior Performance ◽  
Product Reviews ◽  
Review Helpfulness ◽  
Personality Theories ◽  
Deep Learning Model ◽  
Practical Implications

This work examines the question of who is more likely to provide future helpful reviews in the context of online product reviews by synergistically using personality theories and data analytics. It trains a deep learning model to infer a reviewer’s personality traits. This enables analyses to reveal the role of personality traits in review helpfulness among a large population of reviewers. We develop hypotheses on how personality traits are associated with review helpfulness, followed by hypotheses testing that confirms that higher review helpfulness is related to higher openness, conscientiousness, extraversion, and agreeableness and to lower emotional stability. These results suggest the appropriateness of using these five personality traits as inputs for developing a model for predicting future review helpfulness. Based on an ensemble model using supervised classification algorithms, we develop a predictive model and demonstrate its superior performance. Theoretical and practical implications are discussed.

Magnetic Nanoparticles: Current Trends and Future Aspects in Diagnostics and Nanomedicine

2019 ◽  
Vol 20 (6) ◽  
pp. 457-472 ◽  
Author(s):  
Naga Veera Srikanth Vallabani ◽  
Sanjay Singh ◽  
Ajay Singh Karakoti
Keyword(s):  
Magnetic Nanoparticles ◽  
Biomedical Applications ◽  
Antimicrobial Agents ◽  
Superparamagnetic Iron Oxide ◽  
Disease Diagnosis ◽  
Imaging Modalities ◽  
Clinical Settings ◽  
Current Trends ◽  
Recent Developments ◽  
Near Future

Background: Biomedical applications of Magnetic Nanoparticles (MNPs) are creating a major impact on disease diagnosis and nanomedicine or a combined platform called theranostics. A significant progress has been made to engineer novel and hybrid MNPs for their multifunctional modalities such as imaging, biosensors, chemotherapeutic or photothermal and antimicrobial agents. MNPs are successfully applied in biomedical applications due to their unique and tunable properties such as superparamagnetism, stability, and biocompatibility. Approval of ferumoxytol (feraheme) for MRI and the fact that several Superparamagnetic Iron Oxide Nanoparticles (SPIONs) are currently undergoing clinical trials have paved a path for future MNPs formulations. Intensive research is being carried out in designing and developing novel nanohybrids for multiple applications in nanomedicine. Objective: The objective of the present review is to summarize recent developments of MNPs in imaging modalities like MRI, CT, PET and PA, biosensors and nanomedicine including their role in targeting and drug delivery. Relevant theory and examples of the use of MNPs in these applications have been cited and discussed to create a thorough understanding of the developments in this field. Conclusion: MNPs have found widespread use as contrast agents in imaging modalities, as tools for bio-sensing, and as therapeutic and theranostics agents. Multiple formulations of MNPs are in clinical testing and may be accepted in clinical settings in near future.

scholarly journals Masked Label Prediction: Unified Message Passing Model for Semi-Supervised Classification

2021 ◽  
Author(s):  
Yunsheng Shi ◽  
Zhengjie Huang ◽  
Shikun Feng ◽  
Hui Zhong ◽  
Wenjing Wang ◽  
...  
Keyword(s):  
Neural Network ◽  
Message Passing ◽  
Supervised Classification ◽  
State Of The Art ◽  
Label Propagation ◽  
Superior Performance ◽  
Label Prediction ◽  
Label Information ◽  
Message Passing Algorithms ◽  
Prediction Strategy

Graph neural network (GNN) and label propagation algorithm (LPA) are both message passing algorithms, which have achieved superior performance in semi-supervised classification. GNN performs feature propagation by a neural network to make predictions, while LPA uses label propagation across graph adjacency matrix to get results. However, there is still no effective way to directly combine these two kinds of algorithms. To address this issue, we propose a novel Unified Message Passaging Model (UniMP) that can incorporate feature and label propagation at both training and inference time. First, UniMP adopts a Graph Transformer network, taking feature embedding and label embedding as input information for propagation. Second, to train the network without overfitting in self-loop input label information, UniMP introduces a masked label prediction strategy, in which some percentage of input label information are masked at random, and then predicted. UniMP conceptually unifies feature propagation and label propagation and is empirically powerful. It obtains new state-of-the-art semi-supervised classification results in Open Graph Benchmark (OGB).

scholarly journals Optical Imaging Modalities: Principles and Applications in Preclinical Research and Clinical Settings

2020 ◽  
Vol 61 (10) ◽  
pp. 1419-1427 ◽  
Author(s):  
Giacomo Pirovano ◽  
Sheryl Roberts ◽  
Susanne Kossatz ◽  
Thomas Reiner
Keyword(s):  
Optical Imaging ◽  
Imaging Modalities ◽  
Clinical Settings ◽  
Preclinical Research

scholarly journals GET_PHYLOMARKERS, a software package to select optimal orthologous clusters for phylogenomics and inferring pan-genome phylogenies, used for a critical geno-taxonomic revision of the genusStenotrophomonas

2018 ◽  
Author(s):  
Pablo Vinuesa ◽  
Luz Edith Ochoa-Sánchez ◽  
Bruno Contreras-Moreira
Keyword(s):  
Core Genome ◽  
Phylogenetic Analyses ◽  
Superior Performance ◽  
Biological Research ◽  
Species Trees ◽  
Genome Sequences ◽  
Sequence Alignments ◽  
Multiple Sequence ◽  
Pan Genome ◽  
Link Type

AbstractThe massive accumulation of genome-sequences in public databases promoted the proliferation of genome-level phylogenetic analyses in many areas of biological research. However, due to diverse evolutionary and genetic processes, many loci have undesirable properties for phylogenetic reconstruction. These, if undetected, can result in erroneous or biased estimates, particularly when estimating species trees from concatenated datasets. To deal with these problems, we developed GET_PHYLOMARKERS, a pipeline designed to identify high-quality markers to estimate robust genome phylogenies from the orthologous clusters, or the pan-genome matrix (PGM), computed by GET_HOMOLOGUES. In the first context, a set of sequential filters are applied to exclude recombinant alignments and those producing anomalous or poorly resolved trees. Multiple sequence alignments and maximum likelihood (ML) phylogenies are computed in parallel on multi-core computers. A ML species tree is estimated from the concatenated set of top-ranking alignments at the DNA or protein levels, using either FastTree or IQ-TREE (IQT). The latter is used by default due to its superior performance revealed in an extensive benchmark analysis. In addition, parsimony and ML phylogenies can be estimated from the PGM.We demonstrate the practical utility of the software by analyzing 170Stenotrophomonasgenome sequences available in RefSeq and 10 new complete genomes of environmentalS. maltophiliacomplex (Smc) isolates reported herein. A combination of core-genome and PGM analyses was used to revise the molecular systematics of the genus. An unsupervised learning approach that uses a goodness of clustering statistic identified 20 groups within the Smc at a core-genome average nucleotide identity of 95.9% that are perfectly consistent with strongly supported clades on the core- and pan-genome trees. In addition, we identified 14 misclassified RefSeq genome sequences, 12 of them labeled asS. maltophilia, demonstrating the broad utility of the software for phylogenomics and geno-taxonomic studies. The code, a detailed manual and tutorials are freely available for Linux/UNIX servers under the GNU GPLv3 license athttps://github.com/vinuesa/get_phylomarkers. A docker image bundling GET_PHYLOMARKERS with GET_HOMOLOGUES is available athttps://hub.docker.com/r/csicunam/get_homologues/, which can be easily run on any platform.

scholarly journals Precise, Fast and Comprehensive Analysis of Intact Glycopeptides and Monosaccharide-Modifications with pGlyco3

2021 ◽  
Author(s):  
Wen-Feng Zeng ◽  
Wei-Qian Cao ◽  
Ming-Qi Liu ◽  
Si-Min He ◽  
Peng-Yuan Yang
Keyword(s):  
Search Engine ◽  
Software Tools ◽  
Comprehensive Analysis ◽  
The Other ◽  
Identification Accuracy ◽  
Superior Performance ◽  
Isotopic Labelling ◽  
Link Type ◽  
Glycosylation Sites ◽  
Search Speed

AbstractWe presented a glycan-first glycopeptide search engine, pGlyco3, to comprehensively analyze intact N- and O-glycopeptides, including glycopeptides with monosaccharide-modifications. We developed an algorithm, termed pGlycoSite, to localize the glycosylation sites and estimate the localization probabilities. We designed a number of experiments to validate the accuracy of pGlyco3 as well as other frequently used or recently developed software tools. These experiments showed that pGlyco3 outperformed the other tools on both N- and O-glycopeptide identification accuracy especially at the glycan level, without loss of the sensitivity. pGlyco3 also achieved a superior performance in terms of search speed. As pGlyco3 was shown to be accurate and flexible for glycopeptide search with monosaccharide-modifications, we then discovered a monosaccharide-modification of Hex (or an uncommon monosaccharide) “Hex+17.027 Da” on both O-mannose and N-glycopeptides in yeast samples, and confirmed this monosaccharide based on released N-glycans and isotopic labelling data. pGlyco3 is freely available on https://github.com/pFindStudio/pGlyco3/releases.

scholarly journals Riboflow: using deep learning to classify riboswitches with ~99% accuracy

2019 ◽  
Author(s):  
Keshav Aditya R. Premkumar ◽  
Ramit Bharanikumar ◽  
Ashok Palaniappan
Keyword(s):  
Neural Networks ◽  
Deep Learning ◽  
Performance Metrics ◽  
Short Term Memory ◽  
Critical Role ◽  
Predictive Modelling ◽  
Superior Performance ◽  
Genetic Circuits ◽  
Roc Curve Analysis ◽  
Link Type

AbstractRiboswitches are cis-regulatory genetic elements that use an aptamer to control gene expression. Specificity to cognate ligand and diversity of such ligands have expanded the functional repetoire of riboswitches to mediate mounting apt responses to sudden metabolic demands and signal changes in environmental conditions. Given their critical role in microbial life, and novel uses in synthetic biology, riboswitch characterisation remains a challenging computational problem. Here we have addressed the issue with advanced deep learning frameworks, namely convolutional neural networks (CNN), and bidirectional recurrent neural networks (RNN) with Long Short-Term Memory (LSTM). Using a comprehensive dataset of 32 ligand classes and a stratified train-validate-test approach, we demonstrated the superior performance of both the deep models (CNN and RNN) relative to other conventional machine learning classifiers on all key performance metrics, including the ROC curve analysis. In particular, the bidirectional LSTM RNN emerged as the best-performing learning method for identifying the ligand-specificity of riboswitches with an accuracy > 0.99 and macro-averaged F-score of 0.96. A dynamic update functionality is inbuilt to account for the discovery of new riboswitches and extend the predictive modelling to any number of new additional classes. Our work would be valuable in the design and assembly of genetic circuits and the development of the next generation of antibiotics. The software is freely available as a Python package and standalone resource for wide use in genome annotation and biotechnology workflows.AvailabilityPyPi package: riboflow @ https://pypi.org/project/riboflowRepository with Standalone suite of tools: https://github.com/RiboswitchClassifierLanguage: Python 3.6 with numpy, keras, and tensorflow libraries.Licence: MIT

Superior performance of cone-beam CT angiography in characterization of intracranial atherosclerosis

2014 ◽  
Vol 121 (2) ◽  
pp. 441-449 ◽  
Author(s):  
Mina G. Safain ◽  
Jason P. Rahal ◽  
Samir Patel ◽  
Alexandra Lauric ◽  
Edward Feldmann ◽  
...  
Keyword(s):  
Ct Angiography ◽  
Clinical Decision Making ◽  
Cone Beam Ct ◽  
Morphological Characteristics ◽  
Superior Performance ◽  
Imaging Modalities ◽  
Cone Beam ◽  
Plaque Morphology ◽  
Intracranial Atherosclerotic Disease ◽  
The Absolute

Object Intracranial atherosclerotic disease (ICAD) carries a high risk of stroke. Evaluation of ICAD has focused on assessing the absolute degree of stenosis, although plaque morphology has recently demonstrated increasing relevance. The authors provide the first report of the use of ultra-high-resolution C-arm cone-beam CT angiography (CBCT-A) in the evaluation of vessel stenosis as well as plaque morphology. Methods Between August 2009 and July 2012, CBCT-A was used in all patients with ICAD who underwent catheter-based angiography at the authors' institution (n = 18). Lesions were evaluated for maximum degree of stenosis as well as plaque morphological characteristics (ulcerated, calcified, dissected, or spiculated) via digital subtraction angiography (DSA), 3D-rotational angiography (3DRA), and CBCT-A. The different imaging modalities were compared in their assessment of absolute stenosis as well as their ability to resolve different plaque morphologies. Results Lesions were found to have similar degrees of stenosis when utilizing CBCT-A compared with 3DRA, but both 3DRA and CBCT-A differed from DSA in their assessment of the absolute degree of stenosis. CBCT-A provided the most detailed resolution of plaque morphology, identifying a new plaque characteristic in 61% of patients (n = 11) when compared with DSA and 50% (n = 9) when compared with 3DRA. CBCT-A identified all lesion characteristics visualized on DSA and 3DRA. Conclusions CBCT-A provides detailed spatial resolution of plaque morphology and may add to DSA and 3DRA in the evaluation of ICAD. Further prospective study is warranted to determine any benefit CBCTA-A may provide in clinical decision making and risk stratification over existing conventional imaging modalities.

scholarly journals SplitStrains, a tool to identify and separate mixed Mycobacterium tuberculosis infections from WGS data

2021 ◽  
Vol 7 (6) ◽  
Author(s):  
Einar Gabbassov ◽  
Miguel Moreno-Molina ◽  
Iñaki Comas ◽  
Maxwell Libbrecht ◽  
Leonid Chindelevitch
Keyword(s):  
Public Health ◽  
Type Species ◽  
Bacterial Pathogen ◽  
Superior Performance ◽  
Whole Genome Sequencing Data ◽  
Sequencing Data ◽  
Content Type ◽  
Link Type ◽  
Novel Method ◽  
Multiple Strains

The occurrence of multiple strains of a bacterial pathogen such as M. tuberculosis or C. difficile within a single human host, referred to as a mixed infection, has important implications for both healthcare and public health. However, methods for detecting it, and especially determining the proportion and identities of the underlying strains, from WGS (whole-genome sequencing) data, have been limited. In this paper we introduce SplitStrains, a novel method for addressing these challenges. Grounded in a rigorous statistical model, SplitStrains not only demonstrates superior performance in proportion estimation to other existing methods on both simulated as well as real M. tuberculosis data, but also successfully determines the identity of the underlying strains. We conclude that SplitStrains is a powerful addition to the existing toolkit of analytical methods for data coming from bacterial pathogens and holds the promise of enabling previously inaccessible conclusions to be drawn in the realm of public health microbiology.

Sign in / Sign up

Export Citation Format

Share Document