Signals Are All You Need: A Deep Learning Architecture for ECG Lead Classification
A permutation-equivariant deep learning architecture that infers lead labels directly from waveforms, using convolutional, recurrent, and attention components to model temporal and inter-lead structure is proposed.
Nathan Macdonald, R. Peeters, P. Bonizzi et al.
· 0 citations