Validation of a multiple­lead smartphone-based electrocardiograph with automated lead placement for layman use in patients with hypertrophic cardiomyopathy.

ABSTRACT

BACKGROUND: A smartphone 12-Lead ECG that enables layman ECG screening is still lacking. We aimed to validate D-Heart ECG device, a smartphone 8/12 Lead electrocardiograph with an image processing algorithm to guide secure electrode placement by non-professional users. METHODS: One-hundred-fourty-five patients with HCM were enrolled. Two uncovered chest images were acquired using the smartphone camera. An image with virtual electrodes placement by imaging processing algorithm software was compared to the 'gold standard' electrode placement by a doctor. D-Heart 8 and 12-Lead ECG were obtained, immediately followed by 12­lead ECGs and were assessed by 2 independent observers. Burden of ECG abnormalities was defined by a score based on the sum of 9 criteria, identifying four classes of increasing severity. RESULTS: A total of 87(60%) patients presented a normal/mildly abnormal ECG, whereas 58(40%) had moderate or severe ECG alteration. Eight(6%) patients had ≥1 misplaced electrode. D-Heart 8-Lead and 12­lead ECGs concordance according to Cohen's weighted kappa test was 0,948 (p < 0,001, agreement of 97.93%). Concordance was high for the Romhilt-Estes score (kw = 0,912; p < 0.01). Concordance between D-Heart 12-Lead ECG and standard 12-Lead ECG was perfect (kw = 1). PR and QRS intervals measurements comparison with Bland-Altman method showed good accuracy (95% limit of agreement ±18 ms for PR and ± 9 ms for QRS). CONCLUSIONS: D-Heart 8/12-Lead ECGs proved accurate, allowing an assessment of ECG abnormalities comparable to the standard 12­lead ECG in patients with HCM. The image processing algorithm provided accurate electrode placement, standardizing exam quality, potentially opening perspectives for layman ECG screening campaigns.