What question did this study set out to answer?

To develop and evaluate an automated, two-step algorithm that scores electrocardiogram (ECG) recording quality by identifying macroscopic errors and quantifying baseline, powerline, and muscular noise.

August 17, 2012Physiological MeasurementOpen Access

Automatic ECG quality scoring methodology: mimicking human annotators

Q: What is the clinical evidence from this study?

Study design: Other. Population: ECG signal quality (n=1500). Intervention: Improved two-step algorithm for ECG quality scoring vs. Human annotators. Primary outcome: Classification accuracy.

View Full Paper

Ask AI

Bookmark

Key result

An improved two-step algorithm for ECG quality scoring achieved a classification accuracy of 92.3% on the training set and 90.0% on the test set.

Why the study?

Does an improved two-step algorithm accurately score ECG quality compared to human annotators?

Population

1500 ECGs from the PhysioNet Challenge database rated by humans for signal quality

Comparison

Improved two-step algorithm for ECG quality… vs Human annotators (reference standard)

Design

Other

Authors

LJLars JohannesenUnited States Food and Drug Administration LGLoriano GaleottiUnited States Food and Drug Administration

Discussion

Member takes

Implication

May assist ECG recording by inexperienced staff; leaves open prospective clinical validation.

Key Points

To develop and evaluate an automated, two-step algorithm that scores electrocardiogram (ECG) recording quality by identifying macroscopic errors and quantifying baseline, powerline, and muscular noise.
Designed a two-step algorithm that first screens and discards ECGs with major defects (signal absence, large voltage shifts, or saturation) and then calculates a continuous noise score.
Evaluated algorithm performance using the PhysioNet Challenge 2011 database containing 1,500 human-annotated ECG recordings.
The algorithm achieved a classification accuracy of 92.3% on the training dataset.
The classification accuracy reached 90.0% on the independent test dataset while successfully generating continuous overall quality scores.

Structured PICO

Does an improved two-step algorithm accurately score ECG quality compared to human annotators?

Population

1,500 ECGs from the PhysioNet Challenge database rated by humans for signal quality to evaluate an automated quality scoring algorithm.

Intervention

Improved two-step algorithm for ECG quality scoring (rejects macroscopic errors and quantifies noise on a continuous scale)

Comparator

Human annotators (reference standard)

Outcome

Classification accuracysurrogate

An improved two-step algorithm can accurately assess ECG signal quality, potentially aiding inexperienced personnel in recording diagnostic-quality ECGs.

Cite This Study

Johannesen et al. (2012) studied ECG signal quality (n=1,500). Improved two-step algorithm for ECG quality scoring vs. Human annotators was evaluated on Classification accuracy. An improved two-step algorithm for ECG quality scoring achieved a classification accuracy of 92.3% on the training set and 90.0% on the test set.

synapsesocial.com/papers/6a8703baf87adc0a02ca24da https://doi.org/10.1088/0967-3334/33/9/1479

View Full Paper

Ask AI

Bookmark

Automatic ECG quality scoring methodology: mimicking human annotators

Discussion

Member takes

Implication

Key Points

Structured PICO

Cite This Study

Automatic ECG quality scoring methodology: mimicking human annotators

Discussion

Member takes

Implication

Key Points

Structured PICO

Cite This Study

Also Consider

Also Consider