Multitype Learning via Multimodal Data Embedding

2021 ◽  
Author(s):  
Ekrem Yildiz ◽  
Ege Burak Safdil ◽  
Furkan Arslan ◽  
Huseyin Fuat Alsan ◽  
Taner Arsan
Keyword(s):  
Data Embedding ◽  
Multimodal Data

Big Data Driven Clinical Informatics & Surveillance (BDD_CIS) – A Multimodal Database Focused Clinical, Community, and Multi-Omics Surveillance Plan for COVID-19: A study Protocol (Preprint)

2020 ◽  
Author(s):  
Bankole Olatosi ◽  
Jiajia Zhang ◽  
Sharon Weissman ◽  
Zhenlong Li ◽  
Jianjun Hu ◽  
...  
Keyword(s):  
Big Data ◽  
South Carolina ◽  
Data Science ◽  
Age Groups ◽  
The Elderly ◽  
The United States ◽  
Data Sources ◽  
Patient Registries ◽  
Multiple Partner ◽  
Multimodal Data

BACKGROUND The Coronavirus Disease 2019 (COVID-19) caused by the severe acute respiratory syndrome coronavirus (SARS-CoV-2) remains a serious global pandemic. Currently, all age groups are at risk for infection but the elderly and persons with underlying health conditions are at higher risk of severe complications. In the United States (US), the pandemic curve is rapidly changing with over 6,786,352 cases and 199,024 deaths reported. South Carolina (SC) as of 9/21/2020 reported 138,624 cases and 3,212 deaths across the state. OBJECTIVE The growing availability of COVID-19 data provides a basis for deploying Big Data science to leverage multitudinal and multimodal data sources for incremental learning. Doing this requires the acquisition and collation of multiple data sources at the individual and county level. METHODS The population for the comprehensive database comes from statewide COVID-19 testing surveillance data (March 2020- till present) for all SC COVID-19 patients (N≈140,000). This project will 1) connect multiple partner data sources for prediction and intelligence gathering, 2) build a REDCap database that links de-identified multitudinal and multimodal data sources useful for machine learning and deep learning algorithms to enable further studies. Additional data will include hospital based COVID-19 patient registries, Health Sciences South Carolina (HSSC) data, data from the office of Revenue and Fiscal Affairs (RFA), and Area Health Resource Files (AHRF). RESULTS The project was funded as of June 2020 by the National Institutes for Health. CONCLUSIONS The development of such a linked and integrated database will allow for the identification of important predictors of short- and long-term clinical outcomes for SC COVID-19 patients using data science.

A robust and high capacity data hiding method for JPEG compressed images with SVD-based block selection and advanced error correcting techniques

2021 ◽  
pp. 1-11
Author(s):  
Kusan Biswas
Keyword(s):  
Data Hiding ◽  
High Capacity ◽  
Error Correcting Codes ◽  
Human Vision ◽  
Data Embedding ◽  
Compressed Images ◽  
Matrix Encoding ◽  
Dct Coefficients ◽  
Capacity Data ◽  
Value Decomposition

In this paper, we propose a frequency domain data hiding method for the JPEG compressed images. The proposed method embeds data in the DCT coefficients of the selected 8 × 8 blocks. According to the theories of Human Visual Systems  (HVS), human vision is less sensitive to perturbation of pixel values in the uneven areas of the image. In this paper we propose a Singular Value Decomposition based image roughness measure (SVD-IRM) using which we select the coarse 8 × 8 blocks as data embedding destinations. Moreover, to make the embedded data more robust against re-compression attack and error due to transmission over noisy channels, we employ Turbo error correcting codes. The actual data embedding is done using a proposed variant of matrix encoding that is capable of embedding three bits by modifying only one bit in block of seven carrier features. We have carried out experiments to validate the performance and it is found that the proposed method achieves better payload capacity and visual quality and is more robust than some of the recent state-of-the-art methods proposed in the literature.

scholarly journals Commonalities across computational workflows for uncovering explanatory variants in undiagnosed cases

2021 ◽  
Author(s):  
Shilpa Nadimpalli Kobren ◽  
◽  
Dustin Baldridge ◽  
Matt Velinder ◽  
Joel B. Krier ◽  
...  
Keyword(s):  
Online Survey ◽  
Variant Calling ◽  
Theoretical Method ◽  
The United States ◽  
Genomic Sequencing ◽  
Biomedical Data ◽  
Sequencing Data ◽  
Multimodal Data ◽  
Undiagnosed Diseases ◽  
Undiagnosed Diseases Network

Abstract Purpose Genomic sequencing has become an increasingly powerful and relevant tool to be leveraged for the discovery of genetic aberrations underlying rare, Mendelian conditions. Although the computational tools incorporated into diagnostic workflows for this task are continually evolving and improving, we nevertheless sought to investigate commonalities across sequencing processing workflows to reveal consensus and standard practice tools and highlight exploratory analyses where technical and theoretical method improvements would be most impactful. Methods We collected details regarding the computational approaches used by a genetic testing laboratory and 11 clinical research sites in the United States participating in the Undiagnosed Diseases Network via meetings with bioinformaticians, online survey forms, and analyses of internal protocols. Results We found that tools for processing genomic sequencing data can be grouped into four distinct categories. Whereas well-established practices exist for initial variant calling and quality control steps, there is substantial divergence across sites in later stages for variant prioritization and multimodal data integration, demonstrating a diversity of approaches for solving the most mysterious undiagnosed cases. Conclusion The largest differences across diagnostic workflows suggest that advances in structural variant detection, noncoding variant interpretation, and integration of additional biomedical data may be especially promising for solving chronically undiagnosed cases.

scholarly journals Adaptive Model for Biofeedback Data Flows Management in the Design of Interactive Immersive Environments

2021 ◽  
Vol 11 (11) ◽  
pp. 5067
Author(s):  
Paulo Veloso Gomes ◽  
António Marques ◽  
João Donga ◽  
Catarina Sá ◽  
António Correia ◽  
...  
Keyword(s):  
Real Time ◽  
Emotional State ◽  
Adaptive Model ◽  
Multimodal Data ◽  
The Real ◽  
Time Interaction ◽  
Data Flows ◽  
Immersive Systems ◽  
Immersive Environment ◽  
The Relationship

The interactivity of an immersive environment comes up from the relationship that is established between the user and the system. This relationship results in a set of data exchanges between human and technological actors. The real-time biofeedback devices allow to collect in real time the biodata generated by the user during the exhibition. The analysis, processing and conversion of these biodata into multimodal data allows to relate the stimuli with the emotions they trigger. This work describes an adaptive model for biofeedback data flows management used in the design of interactive immersive systems. The use of an affective algorithm allows to identify the types of emotions felt by the user and the respective intensities. The mapping between stimuli and emotions creates a set of biodata that can be used as elements of interaction that will readjust the stimuli generated by the system. The real-time interaction generated by the evolution of the user’s emotional state and the stimuli generated by the system allows him to adapt attitudes and behaviors to the situations he faces.

Sign in / Sign up

Export Citation Format

Share Document