Reliable Deep Learning-Based Detection of Misplaced Chest Electrodes During Electrocardiogram Recording: Algorithm Development and Validation.

Khaled Rjoob,Raymond Bond,Charles Knoery,Aaron Peace,Victoria Mcgilligan,Ali Rababah,Daniel Guldenring,Anne Mcshane,Dewar Finlay,Aleeha Iftikhar,Stephen J Leslie

doi:10.2196/25347

Abstract

BackgroundA 12-lead electrocardiogram (ECG) is the most commonly used method to diagnose patients with cardiovascular diseases. However, there are a number of possible misinterpretations of the ECG that can be caused by several different factors, such as the misplacement of chest electrodes.ObjectiveThe aim of this study is to build advanced algorithms to detect precordial (chest) electrode misplacement.MethodsIn this study, we used traditional machine learning (ML) and deep learning (DL) to autodetect the misplacement of electrodes V1 and V2 using features from the resultant ECG. The algorithms were trained using data extracted from high-resolution body surface potential maps of patients who were diagnosed with myocardial infarction, diagnosed with left ventricular hypertrophy, or a normal ECG.ResultsDL achieved the highest accuracy in this study for detecting V1 and V2 electrode misplacement, with an accuracy of 93.0% (95% CI 91.46-94.53) for misplacement in the second intercostal space. The performance of DL in the second intercostal space was benchmarked with physicians (n=11 and age 47.3 years, SD 15.5) who were experienced in reading ECGs (mean number of ECGs read in the past year 436.54, SD 397.9). Physicians were poor at recognizing chest electrode misplacement on the ECG and achieved a mean accuracy of 60% (95% CI 56.09-63.90), which was significantly poorer than that of DL (P<.001).ConclusionsDL provides the best performance for detecting chest electrode misplacement when compared with the ability of experienced physicians. DL and ML could be used to help flag ECGs that have been incorrectly recorded and flag that the data may be flawed, which could reduce the number of erroneous diagnoses.

Highlights

BackgroundClinicians routinely face the challenge of making sense of a large amount of high-dimensional and heterogeneous data to inform their clinical decision making
A total of 11 medical doctors who were experienced in reading ECGs were recruited to detect electrode misplacement in the second intercostal space (ICS) using the same data set to benchmark the machine learning (ML) and deep learning (DL) models
We can conclude that DL provides the best performance for detecting chest electrode misplacement when compared with ML-based models and the ability of experienced physicians

Summary

Introduction

BackgroundClinicians routinely face the challenge of making sense of a large amount of high-dimensional and heterogeneous data to inform their clinical decision making. Poor clinical decisions can fail to provide the correct diagnosis and treatment, which can have a large impact on patient safety and health care costs [1,2] Artificial intelligence technologies such as deep learning (DL) and machine learning (ML) could play an important role in developing smarter clinical decision-making algorithms that can assist clinicians in making accurate diagnoses. A known error is an incorrectly recorded ECG caused by placing precordial electrodes (chest electrodes: V1, V2, V3, V4, V5, and V6) in incorrect positions, resulting in erroneous ECG signals that are interpreted by physicians to inform patient diagnostic signs and treatment plans. DL and ML could be used to help flag ECGs that have been incorrectly recorded and flag that the data may be flawed, which could reduce the number of erroneous diagnoses

Objectives

Methods

Results

Discussion

Conclusion