---

ECGProcess

ECGProcess is a Python library for reading, processing, and exporting electrocardiogram (ECG) data. It provides a unified interface for extracting waveform signals and clinical metadata from heterogeneous ECG file formats, including XML and DICOM, and mapping them to formats suitable for downstream quality control, statistical analysis, or machine learning pipelines.

Data extraction is driven by user-supplied tab-delimited configuration files that define which signals and metadata fields to retrieve, without requiring changes to library code. The processing pipeline can be readily augmented with user supplied engineering functions, to for example allow for signal resampling or derivation of augmented leads. Processed data can be exported as pandas DataFrames, NumPy arrays, or TensorFlow tf.data.Dataset objects, supporting both in-memory and chunked on-disk workflows.

Additionally, the module provides a flexible plotting API built on top of matplotlib. This can be used for research illustrations as well as to mimic clinical ECG prints. These can be written to disk as illustrations or as numpy arrays (e.g. relevant for imaging based machine learning).

Installation

To install the package from PyPI, run:

pip install ecgprocess

or instead using conda:

conda install afschmidt::ecgprocess

Documentation

Quickstart

• Processing ECG data
• Configuration files to process ECG data

Examples

• Processing XMLs
• Processing DICOMs
• Scaled processing of ECG data
• Visualising ECG signals
• Engineering callables in ECGProcess

Additional documentation

• ECGProcess API

Project admin

• Contributing to the project
• Raising issues

Indices and tables

• Index

• Module Index

• Search Page