Introduction
Pregnancy imposes a substantial physiological burden on the cardiovascular system, often unmasking or exacerbating underlying cardiac conditions. While continuous monitoring is essential for early intervention, the reliance on hospital-grade 12-lead electrocardiograms (ECGs) presents a significant barrier to frequent assessment. Portable ECG solutions offer a promising alternative; however, their performance in the unique hemodynamic environment of pregnancy requires rigorous validation. This study evaluates the "WenXinWuYang," a deep-learning-enhanced, single-lead portable ECG device, specifically within a pregnant cohort.
Methods
A clinical validation study was conducted involving 99 pregnant participants, stratified across all three trimesters. The evaluation protocol consisted of two primary tests:
- Simultaneous Recording: Participants used the portable device while concurrently undergoing a standard 12-lead ECG to establish criterion validity.
- Test-Retest Reliability: A second portable ECG recording was taken within 60 seconds of the first to assess the consistency of the device's measurements.
Key metrics included heart rate (HR), QT interval, QTc interval, and QRS duration. Diagnostic performance for arrhythmia detection was measured against the 12-lead gold standard.
Results
Measurement Accuracy and Agreement
The portable device demonstrated exceptional agreement with the 12-lead ECG. For heart rate, the mean difference was minimal at -0.4 ± 3.1 bpm, with 87.9% of measurements showing strong agreement.
Interval measurements also showed high validity. The QT interval exhibited a mean difference of 2.1 ± 12.0 ms. While QRS duration showed lower correlation (r=0.136), Bland-Altman analysis still indicated acceptable clinical agreement.
Reliability and Arrhythmia Detection
Test-retest reliability was robust, with correlation coefficients of r=0.847 for HR and r=0.836 for QT interval. The device's integrated deep learning model achieved an overall diagnostic accuracy of 94.9% for arrhythmias, with a sensitivity of 84.2% and a specificity of 97.5%.
Discussion
The WenXinWuYang device maintains high accuracy despite the physiological shifts associated with pregnancy, such as changes in heart position and blood volume. Its performance in measuring the QT interval compares favorably to other consumer devices, suggesting its utility as a specialized clinical tool. Limitations include a relatively small sample size and a low prevalence of complex arrhythmias in the study population. Future research should involve larger, more diverse cohorts to further refine the diagnostic algorithms.
Conclusion
The WenXinWuYang portable ECG device is a reliable and valid tool for cardiac monitoring in pregnant women. Its high diagnostic consistency and ease of use make it a viable solution for extending clinical-grade cardiac surveillance into the home environment.
Summary updated