This webinar has been recorded and available on demand.
Clinical trials, both public and proprietary, hold a huge amount of valuable information. Acquiring knowledge from that information in a cost and time efficient manner is a major industry pain point.
Although information from clinical trials is stored in structured or semi-structured form, it is rarely coded with medical terminologies, which creates a significant level of ambiguity and increases the effort for data preparation for analytical purposes.
Therefore our product manager will help you:
- Identify context of medical finding, not just a single concept mentioning
- Formalize free text information adding relations between medical concepts
- Extract and use contextual information for filtering and data partitioning
Great benefits for pharma industry that this knowledge can deliver include:
- Improved discovery and comparison of enriched information
- Efficient retrospective analysis and cross-study analysis
- Еxtracted clinical data to be used in other analytics tasks
You will learn how to apply semantic data normalization to overcome the lack of single reference model (usually implemented as an ontology or a vocabulary) which resolves the ambiguity in natural language. For the demonstration of this innovative approach, we use public clinical studies data from ClinicalTrials.gov. We semantically normalize some of the most important data categories like “Study Condition”, “Study Intervention”, “Adverse Events” and “Population Criteria”. As a result, we identify all major medical concepts and all associated concept qualifiers, which further specify the context in the particular study.