Skip to main content
. 2022 Mar 30;111(9):1010–1017. doi: 10.1007/s00392-022-02012-3

Fig. 1.

Fig. 1

Panels A and B: Illustration of a smartphone-based ECG device (A) with an automated rhythm classification based on traditional machine-learning algorithms (B). Panel C: schematic depiction of a simple neural network designed with one hidden layer. The width of connecting arrows signifies differently weighted connections between layers