analysis pipeline
Recently Published Documents

Document classification is one of the most critical steps in the document analysis pipeline. There are two types of approaches for document classification, known as image-based and multimodal approaches. The image-based document classification approaches are solely based on the inherent visual cues of the document images. In contrast, the multimodal approach co-learns the visual and textual features, and it has proved to be more effective. Nonetheless, these approaches require a huge amount of data. This paper presents a novel approach for document classification that works with a small amount of data and outperforms other approaches. The proposed approach incorporates a hierarchical attention network(HAN) for the textual stream and the EfficientNet-B0 for the image stream. The hierarchical attention network in the textual stream uses the dynamic word embedding through fine-tuned BERT. HAN incorporates both the word level and sentence level features. While the earlier approaches rely on training on a large corpus (RVL-CDIP), we show that our approach works with a small amount of data (Tobacco-3482). To this end, we trained the neural network at Tobacco-3428 from scratch. Thereby, we outperform state-of-the-art by obtaining an accuracy of 90.3%. This results in a relative error reduction rate of 7.9%.

miND is a NGS data analysis pipeline for smallRNA sequencing data. In this protocol, the pipeline is setup and run on an AWS EC2 instance with example data from a public repository. Please see the publication paper on F1000 for more details on the pipeline and how to use it.

The growth of internet usage during the COVID-19 pandemic creates a new business avenue on e-payment for organizations to expand their business horizon. However, challenges on user-related factors arise with this new avenue. This study aims to investigate the association of these factors on the adoption of e-payment services using machine learning inference. An artificial intelligence-based analysis pipeline is established to study the impact of individual items of the dependent factors on the usage of e-payment. In the analysis pipeline, the important items were extracted using a hybrid artificial intelligence method, and the relationships of these items were inferred using the tree algorithm. The results show that items related to expectancy, facilitating conditions, user attitude, and performance expectancy affect usage of e-payment services. Participants below 25 years old require a gamification solution to adopt e-payment, and participants above 40 years old need social support.

Global sea-level rise has been drawing increasingly greater attention in recent years, as it directly impacts the livelihood and sustainable development of humankind. Our research focuses on identifying causal factors and pathways on sea level changes (both global and regional) and subsequently predicting the magnitude of such changes. To this end, we have designed a novel analysis pipeline including three sequential steps: (1) a dynamic structural equation model (dSEM) to identify pathways between the global mean sea level (GMSL) and various predictors, (2) a vector autoregression model (VAR) to quantify the GMSL changes due to the significant relations identified in the first step, and (3) a generalized additive model (GAM) to model the relationship between regional sea level and GMSL. Historical records of GMSL and other variables from 1992 to 2020 were used to calibrate the analysis pipeline. Our results indicate that greenhouse gases, water, and air temperatures, change in Antarctic and Greenland Ice Sheet mass, sea ice, and historical sea level all play a significant role in future sea-level rise. The resulting 95% upper bound of the sea-level projections was combined with a threshold for extreme flooding to map out the extent of sea-level rise in coastal communities using a digital coastal tracker.

Droplet-based experimental platforms allow researchers to perform massive parallelization and high-throughput studies, such as single-cell experiments. Even though there are various options of image analysis software to evaluate the experiment, selecting the right tools require experience and is time consuming. Experts and sophisticated workflow are required to perform the analysis, especially to detect the droplets and analyze their content. There is need for user-friendly droplet analysis pipelines that can be adapted in laboratories with minimum learning curve. Here, we provide a user-friendly workflow for image-based droplet analysis. The workflow comprises of a) CellProfiler-based image-analysis pipeline and b) accompanied with web application that simplifies the analysis and visualization of the droplet-based experiment. We construct necessary modules in CellProfiler (CP) to detect droplets and export the results into our web application. Using the web application, we are able to process and provide basic profiles of the droplet experiment (droplet sizes, droplet signals, sizes-signals plot, and strip plot for each label/condition). We also add a specific module for growth heterogeneity studies in bacteria populations that includes single cell viability analysis and probability distribution of minimum inhibition concentration (MIC) values in population. Our pipeline is usable for both poly- and monodisperse droplet emulsions.

Preclinical studies using rodents have been the choice for many neuroscience researchers due totheir close reflection of human biology. In particular, research involving rodents has utilized MRI to accurately identify brain regions and characteristics by acquiring high resolution cavity images with different contrasts non-invasively, and this has resulted in high reproducibility and throughput. In addition, tractographic analysis using diffusion tensor imaging to obtain information on the neural structure of white matter has emerged as a major methodology in the field of neuroscience due to its contribution in discovering significant correlations between altered neural connections and various neurological and psychiatric diseases. However, unlike image analysis studies with human subjects where a myriad of human image analysis programs and procedures have been thoroughly developed and validated, methods for analyzing rat image data using MRI in preclinical research settings have seen significantly less developed. Therefore, in this study, we present a deterministic tractographic analysis pipeline using the SIGMA atlas for a detailed structural segmentation and structural connectivity analysis of the rat brain’s structural connectivity. In addition, the structural connectivity analysis pipeline presented in this study was preliminarily tested on normal and stroke rat models for initial observation.

Whole-cell modelling is a newly expanding field that has many applications in lab experiment design and predictive drug testing. Although whole-cell model output contains a wealth of information, it is complex and high dimensional and thus hard to interpret. Here, we present an analysis pipeline that combines machine learning, dimensionality reduction, and network analysis to interpret and visualise metabolic reaction fluxes from a set of single gene knockouts simulated in the Mycoplasma genitalium whole-cell model. We found that the reaction behaviours show trends that correlate with phenotypic classes of the simulation output, highlighting particular cellular subsystems that malfunction after gene knockouts. From a graphical representation of the metabolic network, we saw that there is a set of reactions that can be used as markers of a phenotypic class, showing their importance within the network. Our analysis pipeline can support the understanding of the complexity of in silico cells without detailed knowledge of the constituent parts, which can help to understand the effects of gene knockouts and, as whole-cell models become more widely built and used, aid genome design.

We developed an analysis pipeline that can extract microbial sequences from Spatial Transcriptomic data and assign taxonomic labels to them, generating a spatial microbial abundance matrix in addition to the default host expression one, enabling simultaneous analysis of host expression and microbial distribution. We applied it on both human and murine intestinal datasets and validated the spatial microbial abundance information with alternative assays. Finally, we present a few biological insights that can be gained from this novel data. In summary, this proof of concept work demonstrated the feasibility of Spatial Meta-transcriptomic analysis, and pave the way for future experimental optimization.

miND is a NGS data analysis pipeline for smallRNA sequencing data. In this protocol, the pipeline is setup and run on an AWS EC2 instance with example data from a public repository. Please see the publication paper on F1000 for more details on the pipeline and how to use it.

The use of wearable sensors allows continuous recordings of physical activity from participants in free-living or at-home clinical studies. The large amount of data collected demands automatic analysis pipelines to extract gait parameters that can be used as clinical endpoints. We introduce a deep learning-based automatic pipeline for wearables that processes tri-axial accelerometry data and extracts gait events—bout segmentation, initial contact (IC), and final contact (FC)—from a single sensor located at either the lower back (near L5), shin or wrist. The gait events detected are posteriorly used for gait parameter estimation, such as step time, length, and symmetry. We report results from a leave-one-subject-out (LOSO) validation on a pilot study dataset of five participants clinically diagnosed with Parkinson’s disease (PD) and six healthy controls (HC). Participants wore sensors at three body locations and walked on a pressure-sensing walkway to obtain reference gait data. Mean absolute errors (MAE) for the IC events ranged from 22.82 to 33.09 milliseconds (msecs) for the lower back sensor while for the shin and wrist sensors, MAE ranges were 28.56–64.66 and 40.19–72.50 msecs, respectively. For the FC-event detection, MAE ranges were 29.06–48.42, 40.19–72.70 and 36.06–60.18 msecs for the lumbar, wrist and shin sensors, respectively. Intraclass correlation coefficients, ICC(2,k), between the estimated parameters and the reference data resulted in good-to-excellent agreement (ICC ≥ 0.84) for the lumbar and shin sensors, excluding the double support time (ICC = 0.37 lumbar and 0.38 shin) and swing time (ICC = 0.55 lumbar and 0.59 shin). The wrist sensor also showed good agreements, but the ICCs were lower overall than for the other two sensors. Our proposed analysis pipeline has the potential to extract up to 100 gait-related parameters, and we expect our contribution will further support developments in the fields of wearable sensors, digital health, and remote monitoring in clinical trials.