Discovery in Clinical Trials
Gain valuable insights in your clinical research by quickly mining relevant data across various proprietary and public repositories in your role as a pharmaceutical company, clinical trial professional or medical device manufacturer. Leverage semantic technology in combination with AI to extract complex dependencies between adverse events and concomitant medication and more. Model all information using industry standards like MedDRA, SNOMED, CDISC and conform to the highest data protection requirements.
How You’ll Benefit
Leverage smart semantic search
- Save time to find required information in complex clinical documentation;
- Uncover hidden insights by effortlessly mining large number of relevant documents (> 500 000 in our use case).
Adopt FAIR data principles
- Improve data interoperability by applying industry-specific standards (WHODrug, LOINC, CDISC);
- Enrich your proprietary terminologies with public reference vocabularies and ontologies.
Aggregate large volumes of information
- Easily manage information across different systems (eTMF, CTMS), proprietary data and public repositories (PubMED) in one central solution;
- Optimize data quality and regulatory compliance by reliably identifying clinical data discrepancies.
Optimize your data analytics workflows
- Increase productivity by introducing custom automatic workflows for document processing;
- Quickly mine highly specific information from clinical study documents and apply the results in context.
How it Works
Classify, segment and annotate documents
- Classify, segment and annotate documents with multiple text mining pipelines. Different types of entities are extracted.
- Perform text mining on bullet lists, tables and specific document sections.
- Transform all extracted and generated information into RDF. RDF is resolved and made available to other applications.
Filter & classify
Transform all extracted and generated information into RDF. RDF is resolved and made available to other applications
Analyze & contextualize
Index RDF data in SOLRS by preserving its semantic context. Full text search can be performed using highly efficient indexes.
Integrate data extracted from structured and unstructured sources to extend the knowledge base for improved search and discovery.
Speed up research
Allow users to search for specific terms and related concepts to help them get faster insights.
Why It Works
Build and Explore Your Own Knowledge Graph
Both public and proprietary data are integrated into a highly interconnected knowledge graph. The system relies on both semantic instance mappings and advanced information extraction pipelines to identify all semantically related objects among different data sources. As a result, we can create a huge network of semantically interconnected biomedical objects, which could be easily explored using simplified faceted type search interface or just browsing through the objects in the linked data network. The current video explains how the linked life data is built and how the data can be explored.
Drug Dashboards Mash up Information from Many Different Sources
In this video we will demonstrate how we can build comprehensive dashboards, which describe in details a particular category of information. Drug Dashboards mash up information from many different sources and represent it with interactive visualization widgets. One of the most important features is the definition of a Drug Dashboard, which integrates information from many public drug data sources like DrugBank and drugs At FDA with clinical study data from Clinical Trials.gov and Drug Product Brochures from DailyMed.
Enrich Knowledge Graphs Using Document
Ontotext develops advanced information extraction pipelines which are able to categorize documents, recognize semantic sections and semantically annotate content with a wide range of biomedical concept types. Due to the contextualization of the semantic annotations, we can model the extracted data into RDF, which is fused with the structured data already integrated into the knowledge graph of biomedical information. The information extraction pipelines are automated and could process from a single file or web page to a batch of documents.
AstraZaneca used Ontotext technology to develop a platform for Interactive Relationship Discovery that enables the identification of long causal relationship chains between the biomedical objects in the Linked Life Data cloud.