Introduction to electrocardiogram signal quality assessment and estimated accuracy for textile electrodes

Figures

• 

  Figure 1. Annotated PQRST complex.

• 

  Figure 2. Single QRS and PQRST complexes (blue) of a segment aligned on their R peak at \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$T_0$$\end{document}. The individual template complex for each segment (red) was calculated by taking the mean of all individual complexes. The mean Pearson correlation coefficient (r) was calculated based on correlations of single complexes and the template. The QRS and PQRST complexes of the upper (a,b) and lower two figures (c,d) originate from the same segment, respectively.

• 

  Figure 3. Overview of the proposed SQA algorithm. r: average Pearson correlation coefficient between QRS/PQRST complexes and the template within each 15-second segment. std(\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$RR_i$$\end{document}): standard deviation of RR intervals detected within the same 15-second segment.

• 

  Figure 4. Distribution of data across quality classes: (a) for the multi-parameter monitoring belt (Belt) and gel electrodes (Gel) in the inpatient setting (n = 42); (b) the multi-parameter monitoring belt in both outpatient (Home) and inpatient (Hosp) settings (n = 32). Data from Dataset 2. The asterisks indicate significance in the Wilcoxon signed rank test: *p < 0.05, **p < 0.01, ***p < 0.001.

• 

  Figure 5. Relationship between body position and signal quality: (a) Multi-parameter monitoring belt in the inpatient setting; (b) Multi-parameter monitoring belt in the outpatient setting. Data represent the median cumulative proportion of data for all subjects across quality classes: Class 1 (proportion of Class 1), Class 2 (cumulative proportion of Classes 1 and 2), Class 3 (cumulative proportion of Classes 1 to 3), and Class 4 (cumulative proportion across all classes). Data from Dataset 2.

• 

  Figure 6. Influence of body movement on signal quality - relationship between jerk and the proportion data in quality classes. Data from Dataset 2.

• 

  Figure 7. Influence of ECG monitoring belt electrode usage on signal quality. Data from Dataset 1.