[CS 598] Add LUDB Dataset
Type: Dataset Contribution Net IDs: lketema2, in group with (kjkelly2, alya2) Paper: Benchmarking ECG Delineation using Deep Neural Network-based Semantic Segmentation Models Paper Link: https://proceedings.mlr.press/v287/park25a.html Dataset Referenced by paper: https://physionet.org/files/ludb/1.0.1/
Description: Adds the Lobachevsky University Electrocardiography Database (LUDB) dataset implementation. LUDB dataset consists of 200 ECG recordings with 12-lead annotations for P wave, QRS complex, and T wave boundaries. The implementation includes a LUDBDataset class inheriting from BaseDataset with support for WFDB-compatible file formats, automatic metadata generation, parsing utilities for signals and annotations, unit tests, and documentation.
Files added:
pyhealth/datasets/ludb.py
pyhealth/datasets/configs/ludb.yaml
tests/core/test_ludb.py
docs/api/datasets/pyhealth.datasets.LUDBDataset.rst
Files changed:
pyhealth/datasets/__init__.py
docs/api/datasets.rst
pyproject.toml # add wfdb