AF Classification from a Short Single Lead ECG Recording: the Physionet Computing in Cardiology Challenge 2017
The PhysioNet/Computing in Cardiology (CinC) Challenge 2017 focused on differentiating AF from noise, normal or other rhythms in short term (from 9-61 s) ECG recordings performed by patients. A total of 12, 186 ECGs were used: 8, 528 in the public training set and 3, 658 in the private hidden test s...
| Main Authors: | , , , , |
|---|---|
| Other Authors: | |
| Format: | Article |
| Language: | English |
| Published: |
Computing in Cardiology
2020
|
| Online Access: | https://hdl.handle.net/1721.1/126563 |
| _version_ | 1826214205202628608 |
|---|---|
| author | Moody, Benjamin Edward Lehman, Li-Wei Silva, Ikaro Johnson, A. Mark, Roger G |
| author2 | Massachusetts Institute of Technology. Institute for Medical Engineering & Science |
| author_facet | Massachusetts Institute of Technology. Institute for Medical Engineering & Science Moody, Benjamin Edward Lehman, Li-Wei Silva, Ikaro Johnson, A. Mark, Roger G |
| author_sort | Moody, Benjamin Edward |
| collection | MIT |
| description | The PhysioNet/Computing in Cardiology (CinC) Challenge 2017 focused on differentiating AF from noise, normal or other rhythms in short term (from 9-61 s) ECG recordings performed by patients. A total of 12, 186 ECGs were used: 8, 528 in the public training set and 3, 658 in the private hidden test set. Due to the high degree of interexpert disagreement between a significant fraction of the expert labels we implemented a mid-competition bootstrap approach to expert relabeling of the data, levering the best performing Challenge entrants' algorithms to identify contentious labels. A total of 75 independent teams entered the Challenge using a variety of traditional and novel methods, ranging from random forests to a deep learning approach applied to the raw data in the spectral domain. Four teams won the Challenge with an equal high F1 score (averaged across all classes) of 0.83, although the top 11 algorithms scored within 2% of this. A combination of 45 algorithms identified using LASSO achieved an F1 of 0.87, indicating that a voting approach can boost performance. |
| first_indexed | 2024-09-23T16:01:46Z |
| format | Article |
| id | mit-1721.1/126563 |
| institution | Massachusetts Institute of Technology |
| language | English |
| last_indexed | 2024-09-23T16:01:46Z |
| publishDate | 2020 |
| publisher | Computing in Cardiology |
| record_format | dspace |
| spelling | mit-1721.1/1265632022-10-02T05:48:23Z AF Classification from a Short Single Lead ECG Recording: the Physionet Computing in Cardiology Challenge 2017 Moody, Benjamin Edward Lehman, Li-Wei Silva, Ikaro Johnson, A. Mark, Roger G Massachusetts Institute of Technology. Institute for Medical Engineering & Science The PhysioNet/Computing in Cardiology (CinC) Challenge 2017 focused on differentiating AF from noise, normal or other rhythms in short term (from 9-61 s) ECG recordings performed by patients. A total of 12, 186 ECGs were used: 8, 528 in the public training set and 3, 658 in the private hidden test set. Due to the high degree of interexpert disagreement between a significant fraction of the expert labels we implemented a mid-competition bootstrap approach to expert relabeling of the data, levering the best performing Challenge entrants' algorithms to identify contentious labels. A total of 75 independent teams entered the Challenge using a variety of traditional and novel methods, ranging from random forests to a deep learning approach applied to the raw data in the spectral domain. Four teams won the Challenge with an equal high F1 score (averaged across all classes) of 0.83, although the top 11 algorithms scored within 2% of this. A combination of 45 algorithms identified using LASSO achieved an F1 of 0.87, indicating that a voting approach can boost performance. National Institutes of Health (U.S.) (Grant R01-GM104987) 2020-08-13T17:03:28Z 2020-08-13T17:03:28Z 2017-09 2019-10-09T15:36:25Z Article http://purl.org/eprint/type/JournalArticle 0276-6574 2325-887X https://hdl.handle.net/1721.1/126563 Clifford, Gari D. et al. “AF Classification from a Short Single Lead ECG Recording: the Physionet Computing in Cardiology Challenge 2017.” Computing in cardiology, vol. 44, 2017 © 2017 The Author(s) en 10.22489/CINC.2017.065-469 Computing in cardiology Creative Commons Attribution-Noncommercial-Share Alike http://creativecommons.org/licenses/by-nc-sa/4.0/ application/pdf Computing in Cardiology PMC |
| spellingShingle | Moody, Benjamin Edward Lehman, Li-Wei Silva, Ikaro Johnson, A. Mark, Roger G AF Classification from a Short Single Lead ECG Recording: the Physionet Computing in Cardiology Challenge 2017 |
| title | AF Classification from a Short Single Lead ECG Recording: the Physionet Computing in Cardiology Challenge 2017 |
| title_full | AF Classification from a Short Single Lead ECG Recording: the Physionet Computing in Cardiology Challenge 2017 |
| title_fullStr | AF Classification from a Short Single Lead ECG Recording: the Physionet Computing in Cardiology Challenge 2017 |
| title_full_unstemmed | AF Classification from a Short Single Lead ECG Recording: the Physionet Computing in Cardiology Challenge 2017 |
| title_short | AF Classification from a Short Single Lead ECG Recording: the Physionet Computing in Cardiology Challenge 2017 |
| title_sort | af classification from a short single lead ecg recording the physionet computing in cardiology challenge 2017 |
| url | https://hdl.handle.net/1721.1/126563 |
| work_keys_str_mv | AT moodybenjaminedward afclassificationfromashortsingleleadecgrecordingthephysionetcomputingincardiologychallenge2017 AT lehmanliwei afclassificationfromashortsingleleadecgrecordingthephysionetcomputingincardiologychallenge2017 AT silvaikaro afclassificationfromashortsingleleadecgrecordingthephysionetcomputingincardiologychallenge2017 AT johnsona afclassificationfromashortsingleleadecgrecordingthephysionetcomputingincardiologychallenge2017 AT markrogerg afclassificationfromashortsingleleadecgrecordingthephysionetcomputingincardiologychallenge2017 |