Today up to 70% of medical notes in electronic health records (EHR) are stored in text form. Therefore, these records cannot be processed and used for machine data analysis.

The Webiomed.NLP service offers automatically extract features from medical text records. Features include symptoms, blood pressure data, patient height and weight, lab test results and much more.

Product value

3 thouth. supported features

Objective data, symptoms, examination results and much more

Accuracy up to 98%

High accuracy of data extraction

Complicated signs

The system is able to identify signs consisting of several words

Offline use

It can be used as an embedded solution

We help you to use EHR effectively

Webiomed.NLP service is trained to extract the following information from unstructured text records, medical protocols and notes:

  •      Disease symptoms
  •      Objective data of health examinations
  •      Laboratory tests
  •      Results of clinical investigations

The system supports the extraction of 2 931 different features.

How does it work?

  • Doctors make records in the EHR in their own system. It is not necessary to input exra information or close screen forms.
  • The medical information system disables records from the EHR, generates structured electronic medical documents (SEMDs) from them in the CDA 2.0 format and sends data packets to Webiomed.
  • The platform takes each medical document using the Webiomed.NLP service.
  • Then each sign is analyzed by subject errors. Questionable entries are marked as "junk" and are not taken into account in machine learning processing.

What for

  • Today, up to 80% of medical records in EHR are found in text form. We help you choose from available records for data analysis
  • NLP makes it possible to save physicians from repeated "hacking" of the same information - Webiomed provides the necessary data for use without a person
  • The doctor can share time to communicate with the patient.