Detection of Noise Type in Electrocardiogram
Bibliographic record
Abstract
Physicians use electrocardiogram (ECG) to diagnose cardiovascular diseases. It is mainly used in hospital environment; however, with advancements in ambulatory ECG, it is now available outside of hospital environment. Ambulation can lead to contamination of ECG with various noises leading to signal corruption, misdiagnosis, or false alarms. Removal of noise from ECG is possible; however, blindly applying noise removal techniques may reduce the fidelity of the ECG. As such, identification of the noise in the ECG and applying targeted techniques minimize information loss. In this study, a machine learning approach is used to identify the type of noise in ECG. ECG from Physionet's Normal S inus Rhythm Database was contaminated with noise (baseline wandering, electrode motion, and electromyography) from Physionet's MIT-BIH Noise S tress Test Database at different levels and combinations. The chosen machine learning algorithm was Random Forest with 1024 estimators. The Random Forest had a precision and recall of 1.0 when identifying clean ECG. The average precision and recall were 0.47 and 0.63, respectively, for segments with a single type of noise. The average precision and recall were were 0.44 and 0.27, respectively, for segments with multiple types of noise. The drop in precision and recall was due to the misclassification of the ECG with multiple noises as ECG with a single noise; as an example, classification of an ECG with baseline wandering and electromyography as an ECG with baseline wandering. The classifier performed well at identifying any of the noises in segments with multiple types of noises with an average precision and recall of 0.81 and 0.70, respectively. The classifier generally performed well in identifying types of noise in ECG allowing for future work in developing a framework for identification and mitigation of noise.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.009 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".