CASE STUDY

Understand text
in medical journals

A model based approach for understanding medical journals helps automation.

Background

There are nearly 2 millions scientific papers published each year.
Many medical journals contain important suggestions on procedures, precautions etc. which can provide actionable intelligence.

The client in Germany was focused on the automation of information processing from medical journals.

The client had attempted multiple NLP (Natural language processing) approaches to automate the processing of journals, but was not getting the required quality.
The client engaged us to get improve the quality of text processing and create a method of model creation that could be extended in the future.

Solution

The client had some previous NLP results for reference and was happy with the new results from the medical model.

Solution highlights:

  •   Model allowed structured deduction of medical conditions and action to be taken under those conditions
  •   Extensible architecture - This allowed extending the NLP solution for other medical conditions and actions

Results

Based on our medical model and methodology, the client was able to :

  •   Automate gathering of actionable intelligence from medical journals
  •   Apply the actionable intelligence to improve medical services