Ask for White Paper

Contact information

Professional information

Additional information

privacy policy

Savana Research

Automatic extraction of all variables for a clinical study

Try Savana Research now


  • Automatically extracts every variable for all treated patients, filtering by desired search criteria
  • Allows for inference of previously unknown correlations
  • Immediately recruits patients for clinical trials for whom there is no record


Accelerates the generation of medical knowledge, saving months of work on information extraction

  • Automatic variable extraction

    Automated extractionof variable

  • Correlations discovery

    Identificationof correlations

How Savana Research is helping

Real examples from Savana users:

  • Finding patients with history of thyroid cancer
  • Identifying all cases of “triple negative” breast cancer
  • Clustering patients depending on their response to a biologic immunosuppressant
  • Tracing the “patient journey” in multiple myeloma
  • Finding iatrogenic prostatitis produced by prostate biopsies
  • Determining the national prevalence of a given respiratory disease
Training with millions of care episodes allows all variables to be extracted, considering how clinical concepts are grouped in medical documents. For example, it automatically learns, without specific rules, that antibiotics tend to form a conceptual family. This permits the exponential acceleration of variable extraction and suggestion of new ones that may be related to them
  • Combine primary and specialized care

    Combines primary and specialized care

  • Text and analytic data

    Text and analytic data

  • Structured and unstructured information

    Structured and unstructured information

See our success stories

Discover the benefits of using Savana

See success stories

Are you interested in Savana Research?


Contact information

Professional information

Additional information

privacy policy

Request our White Paper now

Request White Paper