Extracting cancer insights from pathology reports using advanced NLP

September 1, 2015

Extraction of cancer-related parameters from pathology reports is a critical but largely manual process performed by researchers and cancer registrars.

Such data is critical for annotating clinical specimens, supporting cohort selection and providing data for cancer registries and population health analytics. This webinar will demonstrate how key parameters such as histology grade, type and category, tumour size, cancer stage, margin assessment and TNM stage can be extracted using Linguamatics I2E. In addition, we will look at disease specific endpoints such as HER2, ER and PR biomarkers.


Simon Beaulah, Director Healthcare Strategy, Linguamatics 
Sharon Zhang, Healthcare Application Specialist, Linguamatics


Next dates to be announced.


35 minutes including Q&A

Request the recording

During the session you will learn how to:

  • Use I2E to extract cancer parameters by exploring the language used in local pathology reports and devising a query strategy.
  • Extract numerical values such as excision margin and tumor size from the text.
  • Analyse and extract node assessment values.
  • Manage cross references between specimen definition and results.

Who should attend?

Cancer centers:

  • Cancer registry
  • Biobanking/clinical annotation
  • Bioinformatics
  • Population health analysts

Medical research:

  • Comparative effectiveness
  • Outcomes analysis
  • Predictive modelling

Pharma, Biotech and related industries interested in:

  • Real-world data access
  • Comparative effectiveness