medspacy
A clinical NLP library built on spaCy for text processing tasks.
Process clinical text, extract concepts, and detect semantic modifiers using Python.
medspacy is a modular Python library built on spaCy designed for clinical text processing and concept extraction. It suits developers and clinical data scientists building custom NLP pipelines for unstructured health records.
What it is
It brings together specialized modules for sentence segmentation, contextual modifier analysis via the ConText algorithm, and clinical section detection. You can preprocess text, extract UMLS concepts through a QuickUMLS fork, and convert unstructured notes into structured data.
Who it's for
It suits Python developers and clinical researchers needing granular control over text pipelines. It does not suit non-technical users or those looking for a turnkey, out-of-the-box SaaS application.
Key features
Modular Architecture
Each component like context analysis and section detection can be used independently within a spaCy pipeline.
ConText Algorithm
Detects semantic modifiers, negation, and uncertainty for extracted clinical concepts.
UMLS Integration
Includes concept extraction compatible with spaCy via a dedicated QuickUMLS fork.
Multi-language Support
Offers varying levels of rule support for languages including French, Dutch, Spanish, and others.
Up and running in 3 steps
Our honest verdict
What's great
- Built natively on top of the widely adopted spaCy framework
- Modularized components allow selective pipeline construction
- Includes implementations for established clinical algorithms like ConText
- Converts processed unstructured text directly into structured database formats
Where it falls short
- Requires solid Python and spaCy proficiency to configure and run
- Non-English language support is partial and varies heavily by language
- Requires manual rule definition for target concepts and domains
Pricing & license
It is free and open source under the MIT license, with no paid hosted option provided by the maintainers; your only cost is running and maintaining it yourself.
Open remote AI-training roles
874 open remote AI-training roles, updated November 15, 2026; most pay $20–$170 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
Use it, or have us set it up
FAQ
Does medspacy support non-English clinical texts?
Yes, but language support varies significantly. While English has full rule support, other languages like French or Dutch only have partial or experimental rule sets.
Which version of spaCy is required?
medspacy requires spaCy v3 and works with compatible versions up to v3.8.2.
Can I extract UMLS concepts with medspacy?
Yes, via the included SpacyQuickUMLS component which is compatible with spaCy and medspacy.
