Medimetry

Reference implementations of published clinical formulas and scores, written in pure Python for software developers and researchers.

Intended use

medimetry reproduces formulas and scoring systems exactly as published in the cited literature (every function lists its primary source). It returns the raw numeric result or the category defined by that publication and nothing else: no interpretation, no recommendation, no clinical guidance.

medimetry is not a medical device. It is not intended for diagnosis, prevention, monitoring, prediction, prognosis, treatment or alleviation of disease, and it must not be used as a basis for clinical decisions. Anyone integrating it into a product with such a purpose is the manufacturer of that product and is solely responsible for its qualification, validation and regulatory compliance (e.g. as software of unknown provenance under IEC 62304). See DISCLAIMER.md.

The implementations are unit-tested against the published point tables and worked examples, but no clinical validation has been performed and no warranty of correctness is given.

• Free software: MIT license

Installation

pip install medimetry

You can also install the in-development version with::

pip install https://github.com/nerdocs/medimetry/archive/refs/heads/main.zip

Growth charts and WHO reference tables

medimetry.growth computes pediatric z-scores/percentiles from LMS tables. The CDC 2000 tables (public domain) are bundled. The WHO Child Growth Standards / Growth Reference 2007 are licensed CC BY-NC-SA 3.0 IGO and cannot be redistributed in an MIT package, so download them yourself once:

from pathlib import Path
from medimetry.growth import download_who_tables
download_who_tables(Path("~/.cache/medimetry/who").expanduser())

Then pass reference=GrowthReference.WHO, data_dir=... to growth_zscore / growth_percentile. See docs/growth.md for indicators, units and limitations.

Development

See Contributing.

Medical Tests Reference Table