Position-wise comparison of handheld 6-lead ECG versus the standard 12-lead ECG in patients with arrhythmia: comparative study

Fig. 1. Simultaneous acquisition and synchronous pairing of 12-lead and 6-lead ECGs. A Illustrations of the standard acquisition positions for each device: 12-lead ECG in the supine position and 6-lead ECG in the sitting position. In the supine position, the 6-lead device was held with both hands, and the left leg electrode was simulated by attaching an additional electrode to the left ankle (highlighted in red). B In this study, to account for positional effects that may influence signal morphology, both 12-lead and 6-lead ECGs were recorded in both supine and sitting positions. Although the HATIV device records a continuous 30-s 6-lead ECG, only the 10-s segment temporally aligned with the 10-s 12-lead ECG was used for synchronous analysis

Fig. 2. Representative examples of beat labelling on lead II for ECG parameter annotation. Paired examples of 12-lead and 6-lead ECGs in sinus rhythm and atrial fibrillation are shown. Each 10-s ECG recording was independently reviewed by a rater for rhythm classification. Three representative beats are selected and annotated. The horizontal purple line represents the isoelectric baseline. In both cases, the three beats were first labelled on the 12-lead ECG, and the corresponding temporally aligned beats (marked with blue dots) were subsequently labelled on the 6-lead ECG. ECG, electrocardiogram

Fig. 3. Study population and data flowchart. A total of 235 paired 12-lead and 6-lead ECGs were obtained from 130 patients, including 115 pairs in the supine position and 120 pairs in the sitting position. None of the patients underwent multiple recordings in the same position. For exploratory analysis, 12-lead ECGs acquired in the supine position were matched post hoc with 6-lead ECGs acquired in the sitting position to evaluate the postural effects. An “ungradable ECG” was defined as a recording in which rhythm interpretation was not possible. In contrast, “noisy recordings” refer to ECGs in which reliable beat-level annotation (e.g., identification of fiducial points) was not feasible, and these were excluded only from the numerical agreement analysis. ECG, electrocardiogram

Fig. 4. Bland–Altman plots comparing ECG parameters between 12-lead and 6-lead ECG. Negative values indicated higher measurements from the 6-lead ECG. LOA was calculated using the mean difference and SD. The upper LOA was defined as the mean difference plus 1.96*SD, and the lower LOA was defined as the mean difference minus 1.96*SD. Cases without identifiable P waves were excluded from PR interval and P amplitude analyses. ECG, electrocardiogram; LOA, limits of agreement; SD, standard deviation; ms, milliseconds; mV, millivolts