Attendees from multiple pharma companies presented valuable new use cases on how they are using Linguamatics I2E’s Natural Language Processing (NLP)-based AI technology to solve big data challenges from bench to bedside – mining unstructured real world data for rapid reporting of patient trends; discovery of new therapeutic indications of drug targets; developing novel biologics; and supporting risk management and drug safety.
Presenters from healthcare shared how they unlock insights by mining Electronic Health Records (EHRs) in some of the most innovative areas in healthcare today - including real world evidence for clinical outcomes; streamlining prior authorization and medical review workflows; and identifying clinical care gaps.
Life Science and Healthcare NLP use case highlights:
- Agios Pharmaceuticals discussed their use of I2E to build libraries of tool compounds in chemical genetic screens. Using this approach with I2E, Agios has been able to bring candidates to late-stage clinical development in half the time it would normally take.
- Atrius Health showed how I2E is being used to support the transition to accountable care, with use cases in identification of high-risk patients, reducing manual chart review for regulatory quality measures and improving clinical documentation.
- Bristol-Myers Squibb explained how I2E allows them to find clinical trials associated with a target class or specific targets rapidly and effectively, using GPCRs as an example.
- Eli Lilly described their use of I2E to mine ClinicalTrials.gov for early phase trials of combination therapies; a search that’s very time-consuming and costly to do manually. Manual review found 7 trials of interest for a given therapeutic area; using I2E they found over 300.
- GlaxoSmithKline shared their data-driven approach to risk management. GSK consolidates internal and external regulatory data feeds and structures the key data using I2E. This approach enables teams to address emerging concerns in a timelier manner, and has accelerated pipeline decisions.
- Mercy spoke about their use of I2E to mine insights from 100,000 heart failure patients and delivering previously unseen RWE insights in just three months (compared to years of effort with conventional approaches), demonstrating how highly accurate outcomes data can be effectively and efficiently extracted from EHRs.
- Novo Nordisk’s team presented their cloud-based system that uses I2E to extract data from medical scientific liaison (MSL) reports, customer call feeds, and other real world data. The Novo Nordisk team shared their workflow, complete with visual dashboards, which enable product brand owners to respond rapidly to patient issues and trends while reducing manual effort.
- Regeneron reported on providing insights to drug discovery teams on GPCR drugs in clinical trials, and reasons for discontinued trials. Applying I2E’s capabilities leads to rapid identification of crucial insights based on large amounts of text data.
- Sanofi presented three different target discovery workflows enabling their drug discovery teams to discover hidden linkages between novel targets and indications, particularly for rare diseases.
- Secure Exchange Solutions introduced SES SPOT, a joint product with Linguamatics providing dramatically increased efficiencies in the approval of treatments, procedures and medical devices between providers and payers. SES SPOT is an AI enhanced workflow to assess a patient’s suitability and provide the reasoning and audit trail to perform claims handling faster and more systematically, and support better patient outcomes.
Linguamatics also shared a number of exciting NLP product updates:
New I2E features released in the last year, including a Documentum connector, increases in query power and better indexing of mixed documents such as PowerPoint, Word, Excel etc. Queries can now be created in an easy-to-maintain modular fashion, supporting greater sharing and re-use, and query quality assessment is now easier with a built-in gold standard evaluation capability.
- There are now over 180 million documents being served on our I2E OnDemand platform with comprehensive coverage of many different ontologies and built in query resources.
- Linguamatics is now an AWS Technology Partner. AWS is used for both I2E OnDemand and iScite with excellent performance results: tens of millions of documents can be processed per week in real-time on a single AWS instance.
- New Web Portal interfaces highlight the flexibility of the I2E Portal Framework to fit the needs of any group in your organization, regardless of text mining expertise, and access to both internal and external sources.
We would love to hear about your organization’s NLP text mining use cases – contact us to share how NLP is helping (or could help) your organization derive more value from your unstructured data - or to learn more about the use cases above.
This schematic demonstrates that right round the circle from bench to bedside, from drug discovery through development and into delivery of healthcare, and back to bench, insight is needed at every stage to answer business critical questions. As the Text Mining Summit talks showed, to get the complete landscape of information to answer these questions, you need to utilize data from unstructured text. Answers to all these questions will help support business & healthcare decisions; and I2E’s NLP-based text mining platform can help.