Lexigram

About Lexigram

Lexigram provides clinical natural language processing APIs that extract and classify medical concepts from unstructured text such as electronic health record notes. The API identifies entities across categories including Disease, Mental Health, Drug, Procedure, Anatomy, and Other, with sublevel subclassifications for each (for example, distinguishing cardiovascular diseases from symptoms, or classifying drugs by the organ system they act on).

Beyond entity extraction, the API detects structural elements of clinical notes, including document sections (such as "Chief Complaint") and lists (such as numbered problem or medication lists), and aligns matched concepts to the sections and list items in which they appear via character position boundaries. It also extracts and normalizes patient demographic data, including gender, age, ethnicity, and race, by recognizing clustered demographic statements in text.

Access requires an API key obtained through the Lexigram account portal, which is exchanged for an expiring authentication token used with the API endpoints (POST Extract Entities and POST Highlight). The documentation includes an interactive UI demo playground, example code, and a Postman collection for testing without writing code. The service is intended for developers building healthcare applications that need to process clinical text.

Key features

• Extracts and classifies clinical concepts (disease, drug, procedure, anatomy, mental health) from text
• Detects document sections (e.g. Chief Complaint) and aligns matched concepts to their section
• Identifies and parses lists (e.g. problem or medication lists) and links items to matched concepts
• Extracts and normalizes patient demographics such as gender, age, ethnicity, and race
• Detects negation, hypothetical, and other contextual modifiers on extracted concepts
• Provides POST endpoints for entity extraction and text highlighting

Frequently asked questions